Skip to main content

Full text of "ERIC ED459355: A Comparative Analysis of Transitions from Education to Work in Europe (CATEWE). Final Report [and] Annex to the Final Report."

See other formats


DOCUMENT  RESUME 


ED  459  355 


CE  082  741 


AUTHOR 

TITLE 


INSTITUTION 


SPONS  AGENCY 
PUB  DATE 
NOTE 


CONTRACT 
AVAILABLE  FROM 


PUB  TYPE 
EDRS  PRICE 
DESCRIPTORS 


IDENTIFIERS 


Smyth,  Emer;  Gangl,  Markus;  Raffe,  David;  Hannan,  Damian  F.; 
McCoy,  Selina 

A Comparative  Analysis  of  Transitions  from  Education  to  Work 
in  Europe  (CATEWE) . Final  Report  [and]  Annex  to  the  Final 
Report . 

Economic  and  Social  Research  Inst.,  Dublin  (Ireland) . ; 
Economic  and  Social  Research  Council,  Edinburgh  (Scotland) . 
Centre  for  Educational  Sociology.;  Centre  d' Etudes  et  de 
Recherches  sur  les  Qualifications,  Marseilles  (France). 
Commission  of  the  European  Communities,  Brussels  (Belgium) . 
2001-01-30 

776p . ; Prepared  with  Walter  Muller,  Cristina  Iannelli,  Karen 
Brannen,  Maarten  Wolbers,  Michele  Mansuy,  Yvette  Grelet, 
Thomas  Couppie,  Gwenaelle  Thomas,  Hans  Rutjes,  Jannes 
Hartkamp,  and  Lena  Schroder.  Other  partners  include: 
Mannheimer  Zentrum  fur  Europaische  Sozialf orschung  ; 

Research  Centre  for  Education  and  the  Labour  Market, 
Maastricht;  DESAN  Market  Research,  Amsterdam;  and  Instituet 
for  Social  Forskning,  Stockholm.  Proposal  no:  ERB142 
PL97/2100.  Funded  by  the  Targeted  Socio-Economic  Research 
program,  European  Commission  and  by  Deutsche 
Forschungsgemeinschaf t . 

SOE2 -CT97 - 2019 
For  full  text: 

ht tp : / /www . mzes . uni -mannheim . de/pro j ekte/ catewe/publ / publ_e . 
html . 

Reports  - Research  (143) 

MF05/PC32  Plus  Postage. 

Apprenticeships;  Comparative  Analysis;  *Comparative 
Education;  Demography;  *Education  Work  Relationship; 
♦Educational  Attainment;  Educational  Status  Comparison; 
♦Employment  Patterns;  Entry  Workers;  Foreign  Countries; 
Immigrants;  International  Studies;  *Labor  Market;  Outcomes 
of  Education;  Postsecondary  Education;  Secondary  Education; 
Sex  Differences;  Social  Status;  Transitional  Programs; 
Unemployment;  Vocational  Education;  * Young  Adults;  Youth 
♦Europe;  France;  Ireland;  Italy;  Netherlands;  Scotland; 
Spain;  Sweden;  United. Kingdom;  West  Germany 


ABSTRACT 


This  project  aimed  to  develop  a more  comprehensive 
conceptual  framework  of  school -to -work  transitions  in  different  national 
contexts  and  apply  this  framework  to  the  empirical  analysis  of  transition 
processes  across  European  countries.  It  drew  on  these  two  data  sources: 
European  Community  Labor  Force  Survey  and  integrated  databases  on  national 
school  leavers'  surveys  in  France,  Ireland,  the  Netherlands,  Scotland,  and 
Sweden.  Three  broad  types  of  national  systems  were  identified:  countries  with 
extensive  vocational  training  systems  at  upper  secondary  level,  linked  to 
occupational  labor  markets  (Germany,  the  Netherlands) ; countries  with  more 
general  education  systems  with  weaker  institutionalized  linkages  to  the  labor 
market  (Ireland) ; and  Southern  European  (SE)  countries  with  less  vocational 
specialization  and  lower  overall  attainment  than  the  other  groups.  In 


Reproductions  supplied  by  EDRS  are  the  best  that  can  be  made 
from  the  original  document. 


"vocational"  systems,  young  people  tended  to  make  a smoother  transition  into 
the  labor  market,  while  those  in  SE  countries  found  it  more  difficult  to  - 
achieve  a stable  employment  position.  Educational  level  was  highly  predictive 
of  transition  outcomes,  which  varied  by  gender,  social  class,  and  national 
origin.  Early  educational  failure  had  serious  negative  consequences  for  young 
people  across  all  systems.  Sixty-three  references  are  listed.  A separate 
annex  contains  these  17  working  papers:  "Education  and  Unemployment"  (Brauns, 
et  al.);  "Position  of  Young  People  and  New  Entrants  in  European  Labor 
Markets"  (Couppie,  Mansuy) ; "New  Entrants  and  Experienced  Workers  on  European 
Labor  Markets"  (Couppie,  Mansuy) ; "European  Perspectives  on  Labor  Market 
Entry"  (Gangl) ; "Education  and  Labor  Market  Entry  Across  Europe  over  the  Last 
Decade"  (Gangl) ; "Changing  Labor  Markets  and  Early  Career  Outcomes"  (Gangl) ; 
"Transition  from  School  to  Work  in  Southern  Europe"  (Iannelli) ; "Educational 
Attainment  of  Young  People  in  the  European  Union  (EU) " (Mueller,  Wolbers) ; 
"Integration  of  Young  People  into  the  Labor  Market  Within  the  EU"  (van  der 
Velden,  Wolbers) ; "Learning  and  Working"  (Wolbers) ; "Transition  Process" 
(Grelet , et  al.);  "Route  to  Skills"  (Hartkamp,  Rutjes)  ; "Apprenticeship  in 
Ireland,  the  Netherlands,  and  Scotland"  (Hartkamp,  Rutjes) ; "School  Effects 
on  Youth  Transitions  in  Ireland,  Scotland,  and  the  Netherlands"  (Iannelli, 
Soro-Bonmati) ; "Young  Immigrants  on  the  Labor  Market  in  France  and  Sweden" 
(Mansuy,  Schroeder) ; "Relative  Labor  Market  Disadvantage  Among  the  Least 
Qualified  in  Ireland,  the  Netherlands,  and  Scotland,  1979-97"  (McCoy) ; and 
"Gender  Differentiation  in  Education  and  Early  Labor  Market  Transitions" 
(Smyth) . (YLB) 


Reproductions  supplied  by  EDRS  are  the  best  that  can  be  made 
from  the  original  document. 


cn 

m 


s 


A Comparative  Analysis  of  Transitions 
from  Education  to  Work  in  Europe  (CATEWE). 
Final  Report 
[and] 

Annex  to  the  Final  Report 


Emer  Smyth,  Markus  Gangl,  David  Raffe, 
Damian  F.  Hannan,  and  Selina  McCoy 


U.S.  DEPARTMENT  OF  EDUCATION 
Office  of  Educational  Research  and  Improvement 

PERMISSION  TO  REPRODUCE  AND 

EDUCATIONAL  RESOURCES  INFORMATION 

DISSEMINATE  THIS  MATERIAL  HAS 

CENTER  (ERIC) 

BEEN  GRANTED  BY 

□ This  document  has  been  reproduced  as 
received  from  the  person  or  organization 
originating  it. 

E.  Smyth 

□ Minor  changes  have  been  made  to  improve 
reproduction  quality 

TO  THE  EDUCATIONAL  RESOURCES 

• Points  of  view  or  opinions  stated  in  this 

document  do  not  necessarily  represent  official 
OERI  position  or  policy. 

INFORMATION  CENTER  (ERIC) 

1 

rs 

CM 

oo 

o 

LU 

u 


2 


BEST  COPY  AVAILABLE 


FINAL  REPORT 


RESTRICTED 


Contract  no. : SOE2-CT97-20 1 9 

Proposal  no.:  ERB142  PL97/2100 

Title:  A Comparative  Analysis  of  Transitions  from  Education  to 

Work  in  Europe  (CATEWE) 

Project  co-ordinator:  Economic  and  Social  Research  Institute  (ESRI),  Dublin 

Partners:  Centre  for  Educational  Sociology  (CES),  Edinburgh 

Mannheimer  Zentrum  fur  Europaische  Sozialforschung 
(MZES) 

Research  Centre  for  Education  and  the  Labour  Market  (RO  A), 
Maastricht 

Centre  d'Etudes  et  de  Recherches  sur  les  Qualifications 
(CEREQ),  Marseille 
DESAN  Market  Research,  Amsterdam 
Institutet  for  social  forskning  (SOFI),  Stockholm 

Reference  period:  from  December  1997  to  January  2001 

Starting  date:  December  1 997  Duration:  3 years 


Date  of  issue:  30  January  2001 


Project  financed  within  the  TSER  programme 

BEST  COPY  AVAILABLE 


PERMISSION  TO  REPRODUCE  AND 
DISSEMINATE  THIS  MATERIAL  HAS 
BEEN  GRANTED  BY 


8 


TO  THE  EDUCATIONAL  RESOURCES 
INFORMATION  CENTER  (ERIC) 


O 

o 


Off  Sof^du^Uon^l^se^h 

EDUCATIONAL  RESOURCES  INFORMATION 
/ CENTER  (ERIC) 

J This  document  has  been  reproduced  as 
■ received  from  the  person  or  organization 
originating  it. 

□ Minor  changes  have  been  made  to 

improve  reproduction  quality. 


Points  of  view  or  opinions  stated  in  this 
documemdN  not  necessarily  represent 


1 


A Comparative  Analysis  of 

Transitions  from  Education  to  Work  in  Europe  (CATEWE): 

Final  Report 

This  report  was  prepared  by: 

Emer  Smyth  (ESRI) 

Markus  Gangl  (MZES) 

David  Raffe  (CES) 

Damian  F.  Hannan  (ESRI) 

Selina  McCoy  (ESRI) 

With  the  assistance  and  co-operation  of: 

Walter  Muller  (MZES) 

Cristina  Iannelli,  Karen  Brannen  (CES) 

Maarten  Wolbers  (ROA) 

Michele  Mansuy,  Yvette  Grelet,  Thomas  Couppie,  Gwenaelle  Thomas  (CEREQ) 
Hans  Rutjes,  Jannes  Hartkamp  (DESAN) 

Lena  Schroder  (SOFI) 


4 


Table  of  contents 


Abstract i 

1.  Executive  Summary 1 

2.  Background  and  objectives 13 

3.  Scientific  description  of  project  results  and  methodology 18 

4.  Conclusions  and  policy  recommendations 89 

5.  Dissemination 114 

6.  References 120 


Annexes  (in  separate  volume) 

1.  Project  deliverables 

2.  Working  papers 


ABSTRACT 


The  two  main  aims  of  the  CATEWE  project  were  to  develop  a more  comprehensive 
conceptual  framework  of  school  to  work  transitions  in  different  national  contexts,  and 
apply  this  framework  to  the  empirical  analysis  of  transition  processes  across  European 
countries.  The  project  drew  on  two  complementary  data  sources  for  these  analyses: 
the  European  Community  Labour  Force  Survey,  and  integrated  databases  based  on 
national  school  leavers'  surveys  in  France,  Ireland,  the  Netherlands,  Scotland  and 
Sweden. 

Transition  processes  and  outcomes  were  found  to  vary  significantly  across  European 
countries.  Three  broad  types  of  national  system  were  identified:  countries  with 
extensive  vocational  training  systems  at  upper  secondary  level,  linked  to  occupational 
labour  markets  (such  as  Germany  and  the  Netherlands);  countries  with  more  general 
educational  systems  with  weaker  institutionalised  linkages  to  the  labour  market  (such 
as  Ireland);  and  Southern  European  countries  with  less  vocational  specialisation  and 
lower  overall  attainment  than  the  other  groups.  Each  of  these  groups  had  distinctive 
patterns  of  labour  market  integration  among  young  people.  In  'vocational'  systems, 
young  people  tend  to  make  a smoother  transition  into  the  labour  market  while  those  in 
Southern  European  countries  find  it  more  difficult  to  achieve  a stable  employment 
position. 

Across  European  countries,  educational  level  is  highly  predictive  of  transition 
outcomes;  those  with  lower  levels  of  education  have  higher  unemployment  risks  and 
greater  chances  of  entering  low-skilled,  lower  status  and/or  temporary  jobs.  Those 
who  have  taken  part  in  vocational  education/training  (especially  apprenticeships)  tend 
to  have  a smoother  transition  to  their  first  job  and  achieve  more  stable  employment. 
Other  dimensions  of  education  are  also  significant  with  examination  grades  having  a 
greater  effect  in  more  general  education  systems.  Transition  outcomes  are  found  to 
vary  by  gender,  social  class  background  and  national  origin.  There  is  no  evidence  that 
such  differences  have  become  less  important  in  shaping  the  transition  process  over 
time. 

Given  the  diversity  in  education,  training  and  labour  market  systems  across  Europe, 
the  same  policy  interventions  are  unlikely  to  be  equally  effective  in  different  contexts. 
However,  early  educational  failure  has  serious  negative  consequences  for  young 
people  across  all  systems.  There  is  a need,  therefore,  for  policy  intervention  to  reduce 
such  failure  and/or  to  provide  alternative  routes  to  skill  acquisition  for  young  people. 
There  is  also  a need  to  monitor  differences  among  groups  of  young  people  in  terms  of 
gender,  social  class  and  ethnicity  and  pursue  policies  to  address  these  inequalities. 

The  project  highlighted  a number  of  areas  which  should  be  prioritised  in  future 
research:  the  role  of  field  of  education/training  in  transitions,  employer  recruitment 
strategies  in  relation  to  young  people;  young  people's  own  views  of  the  transition 
process;  the  role  of  policy  interventions  (especially  youth  programmes);  and 
regional/local  differences  in  educational  and  transition  outcomes.  It  is  recommended 
that  a European-wide  survey  should  be  initiated,  covering  young  people  from  around 
the  age  of  fifteen  and  following  them  over  a ten-year  period.  In  the  absence  of  such  a 
survey,  we  recommend  that  the  Commission  should  encourage  agreement  on  a 'best 
practice'  template  to  facilitate  the  partial  harmonisation  of  existing  transition  surveys, 
and  use  of,  and  access  to,  the  transitions  module  of  the  Labour  Force  Survey  should 
be  enhanced. 


CHAPTER  1:  EXECUTIVE  SUMMARY 


1.1  PROJECT  BACKGROUND  AND  OBJECTIVES 

Recent  decades  have  seen  rapid  educational  expansion  and  labour  market  changes 
across  European  countries.  Such  changes  have  had  the  greatest  impact  on  those 
entering  the  labour  market  for  the  first  time.  Indeed,  many  commentators  have  argued 
that  the  period  of  transition  from  school  to  work  has  become  more  prolonged  and  less 
predictable  as  a result  of  such  changes.  It  is,  therefore,  crucial  that  we  understand  the 
way  in  which  the  education,  training  and  labour  market  systems  interact  to  shape  the 
transition  process  in  modem  Europe. 

In  spite  of  the  importance  of  the  transition  issue,  existing  research  has  not  yielded  an 
adequate  understanding  of  the  processes  at  work  across  European  countries.  Cross- 
national studies  have  often  focused  on  a narrow  range  of  countries  and  have 
frequently  neglected  important  dimensions  of  the  transition  from  school  to  work.  The 
CATEWE  project  set  out  to  fill  this  gap  by  developing  a more  adequate  conceptual 
framework  to  examine  the  relationship  between  education,  training  and  labour  market 
systems  in  different  national  contexts,  and  by  applying  this  framework  to  empirical 
studies  of  transition  processes  in  several  European  countries. 

The  original  objectives  of  the  project  were: 

1.  To  develop  a more  adequate  and  comprehensive  conceptual  framework,  drawing 
on  existing  research  and  new  analyses,  of: 

• the  nature  of  national  systems  of  initial  education  and  vocational  training  in 
European  societies; 

• the  factors  and  processes  affecting  variation  in  the  full  range  of  education/ 
training  outcomes  by  different  groups  of  young  people  in  each  system; 

• the  processes  of  transition  from  initial  education/training  to  work,  the  main 
outcomes  of  such  transitions  and  pathways,  and  the  main  factors  affecting 
success  and  failure  in  such  transitions  in  each  system; 

• the  impact  of  national  institutional  differences  on  education/training  outcomes 
and  transition  processes  among  young  people. 


2.  Using  this  conceptual  framework,  to  construct  an  integrated  cross-national  data  set 
using  national  school  leavers'  surveys  for  France,  Ireland,  the  Netherlands  and 
Scotland  (UK),  countries  with  widely  varying  education/training  systems  and 
labour  market  structures. 

3.  To  analyse  education  to  work  transitions  across  all  European  countries  using  the 
Labour  Force  Surveys,  placing  the  analysis  of  school  leaver  transitions  in  the 
context  of  the  wider  European  context. 

4.  Using  these  comprehensive  datasets, 

• To  test  and  refine  the  conceptual  framework  to  develop  a more  adequate  and 
comprehensive  framework  to  study  school  to  work  transitions  across  all 
European  countries; 

• To  explore  national  similarities  and  differences  in  ET  systems  and  their 
outcomes,  at  an  aggregate  level  and  for  different  groups  of  young  people;  and 
the  way  in  which  national  differences  in  these  respects  are  influenced  by 
institutional  factors; 

• To  identify  the  main  factors  influencing  success  or  failure  in  ET  outcomes  and 
labour  market  integration  in  each  system;  and  attempt  to  explain  similarities  and 
differences  in  these  patterns  across  the  different  national  systems. 

5.  To  develop  proposals  to  harmonise  existing  school  leavers'  surveys  in  the 
participating  countries;  and  encourage  the  extension  of  comparative  transition 
surveys  to  other  European  countries  currently  planning  surveys  of  school  leavers. 

In  the  course  of  the  project,  these  objectives  were  broadened  in  a number  of  ways. 
Firstly,  analyses  of  the  Eurostat  Labour  Force  survey  indicated  that  the  Southern 
European  countries  had  a distinctive  profile  in  terms  of  labour  market  entry,  an  area 
of  research  that  had  hitherto  been  neglected.  For  this  reason,  it  was  decided  to  further 
explore  the  specific  situation  of  young  people  in  the  Mediterranean  countries  by 
analysing  Labour  Force  Survey  microdata  for  Spain  and  Italy.  Secondly,  in  order  to 
broaden  the  number  of  countries  included  in  the  school  leavers'  survey  database, 


contacts  were  established  with  a Swedish  partner  to  utilise  the  Swedish  cohort  survey 
of  young  people.  The  inclusion  of  Sweden  represented  a substantive  addition  to  the 
project  since  it  has  a number  of  distinctive  characteristics  in  terms  of  the  transition 
process,  including  a long-term  decline  in  inequality  of  educational  opportunity  (on  the 
basis  of  socio-economic  background),  the  absence  of  a formal  apprenticeship  system 
coupled  with  extensive  provision  of  youth  programmes.  Further  exploration  of  the 
availability  and  coverage  of  school  leavers'  survey  data  indicated  the  value  of  going 
beyond  the  original  parameters  of  the  project  to  construct  comparative  databases  not 
only  for  a recent  time-point  but  to  allow  us  to  study  trends  over  time  as  well  as  young 
people's  experiences  over  their  first  five  years  in  the  labour  market.  Due  to  our 
concern  with  contributing  to  the  debate  on  data  harmonisation,  the  CATEWE  team 
initiated  co-operation  with  two  groups  (in  Flanders  and  Portugal)  who  were  planning 
national  transition  surveys. 

1.2  CONCEPTUAL  FRAMEWORK 

The  initial  conceptual  framework  identified  three  sets  of  dimensions  necessary  to 
explore  transitions  in  comparative  perspective:  the  demographic,  economic  and  labour 
market  context  within  which  transitions  occur;  the  dimensions  of  the  education  and 
training  system;  and  the  nature  of  transition  processes  and  outcomes. 

The  context  within  which  the  transition  process  has  often  been  neglected  in 
comparative  studies  of  labour  market  entry.  However,  important  contextual 
differences  are  evident  among  European  countries.  Firstly,  they  differ  not  only  in  the 
industrial  and  occupational  structure  of  employment  but  also  in  relation  to  the  type  of 
labour  market  structuration  (particularly  whether  occupational  or  internal  labour 
markets  have  greater  importance).  Significant  variation  is  also  evident  in  the  extent  of 
labour  market  regulation  which  is  likely  to  affect  the  ease  of  access  of  young  people 
to  (stable)  employment.  Macroeconomic  conditions,  including  the  nature  of  economic 
development  and  the  stage  in  the  economic  cycle,  are  also  likely  to  have  a 
disproportionate  impact  on  young  people's  labour  force  situation.  Other  contextual 
factors  which  need  to  be  considered  include  the  nature  of  the  family-  and  State-based 
welfare  provision,  and  the  age,  gender  and  ethnic  structuring  of  the  labour  force  as  a 
whole. 


0 


3 


9 


Previous  comparative  research  on  transitions  tended  to  focus  on  two  aspects  of 
education/training  systems,  standardisation  (that  is,  the  extent  to  which  curricula, 
assessment  and  certification  are  nationally  or  regionally  standardised)  and  track 
differentiation  (the  division  between  vocational  and  academic/general  tracks).  The 
CATEWE  project  went  beyond  these  approaches  to  examine  a broader  range  of 
dimensions.  In  particular,  we  were  concerned  with  examining  the  way  in  which 
education/training  systems  'sort'  or  differentiate  their  students  not  just  in  terms  of 
programmes  within  a stage  (for  example,  between  vocational  or  academic)  but  also  in 
terms  of  progression  into  the  next  stage  and  outcomes  at  the  end  of  a stage  (such  as 
examination  grades).  We  also  considered  the  nature  of  school  to  work  linkages,  that 
is,  the  role  of  employers  in  the  education/training  system,  as  an  important  dimension. 
Our  framework  did  not  just  focus  on  school-based  education  but  also  explored  cross- 
national variation  in  the  nature  of  youth  training  (through  apprenticeships  and  youth 
programmes),  along  with  its  linkage  to  initial  education  and  the  labour  market. 

These  different  dimensions  of  the  education/training  system  can  be  seen  to  interact 
with  contextual  features  to  produce  a 'transition  system':  the  relatively  enduring 
features  of  a country's  institutional  and  structural  arrangements  which  shape  transition 
processes  and  outcomes.  The  important  features  of  the  transition  process  were  seen  to 
include  the  number  and  sequence  of  transitions  and  the  nature  of  the  resulting 
trajectories,  the  length  of  the  transition  period  (from  leaving  education  to  'stable' 
employment,  for  example),  differentiation  between  transition  statuses  (for  example, 
between  'employment'  and  youth  training),  and  inequalities  in  terms  of  gender,  social 
class  and  ethnicity.  A wide  range  of  transition  outcomes  were  considered  including: 
principal  economic  status;  occupational  status;  industrial  allocation;  labour  market 
segment;  pay;  and  access  to  training  among  young  people. 

While  conceptually  distinct,  different  dimensions  of  the  framework  interact  to 
produce  clusters  of  national  systems.  A broad  continuum  of  European  systems  is 
evident,  ranging  from  countries  with  high  standardisation,  strong  track  differentation 
and  strong  linkages  between  education  and  the  labour  market  (for  example,  Germany 
and  the  Netherlands)  to  countries  with  equally  high  standardisation  but  much  weaker 
track  differentiation  and  school  to  work  linkages  (Ireland  and  Scotland,  albeit  with 


strong  market  signals  in  terms  of  educational  qualifications).  However,  additional 
features  which  may  cross-cut  this  continuum  must  be  considered,  including  the 
strength  of  labour  market  regulation  within  a national  system,  along  with  the  nature  of 
the  formal  and  informal  (primarily  family-based)  welfare  regimes,  a feature  which  has 
particular  relevance  to  Mediterranean  countries.  Finally,  it  should  also  be  noted  that 
different  parts  of  an  education/training  system  may  have  different  characteristics  (for 
example,  school-based  provision  may  differ  markedly  in  nature  from  post-school 
vocational  training). 

1.3  DATA  SOURCES 

The  CATEWE  project  drew  on  two  data  sources  with  complementary  strengths  for 
the  analysis  of  school  to  work  transitions:  the  Eurostat  Labour  Force  Survey  (EULFS) 
and  integrated  databases  based  on  national  school  leavers'  surveys  (SLS). 

The  EULFS  is  a cross-sectional  survey,  covering  all  EU  member  states,  and  is 
constructed  to  a comparable  framework  in  order  to  facilitate  cross-national 
comparison.  As  such,  it  enables  us  to  study  the  full  diversity  of  national  contexts 
across  Europe  and  to  compare  the  situation  of  young  people  with  that  of  their  adult 
counterparts.  It  also  contains  detailed  information  on  employment  outcomes,  allowing 
us  to  examine  their  relationship  with  level  and  type  of  education.  However,  the  cross- 
sectional  nature  of  the  survey  means  that  we  cannot  directly  examine  the  transition 
process  itself.  Instead,  synthetic  labour  force  entry  cohorts  (based  on  'typical'  age  of 
graduation  from  different  levels  and  types  of  education)  have  been  constructed  to 
examine  patterns  across  different  cohorts. 

A set  of  integrated  databases  were  constructed  drawing  on  national  transition  surveys 
in  France,  Ireland,  the  Netherlands,  Scotland  and  Sweden.  Key  measures  of  education 
and  transition  outcomes  were  available  in  these  national  surveys  and  they  were 
mapped  to  a common  template  of  comparable  variables,  a process  which  also  helped 
us  to  refine  the  conceptual  framework  described  above.  The  advantage  of  this  data 
source  is  that  national  school  leavers'  survey  explicitly  take  account  of  the  transition 
process  by  examining  the  flow  out  of  initial  education  into  the  labour  market  or 
further  education/training  and  allow  us  to  observe  individual  pathways  rather  than  the 


aggregate  patterns  possible  with  the  EULFS.  These  surveys  also  contain  detailed 
information  on  the  main  dimensions  of  education/training  outcomes,  enabling  us  to 
explore  their  relative  importance  in  different  institutional  contexts.  However,  the  use 
of  integrated  databases  is  somewhat  limited  in  terms  of  the  number  of  European 
countries  covered,  the  absence  of  information  on  older  age-groups  for  the  purposes  of 
comparison,  and  some  difficulties  in  comparability  (for  example,  the  absence  of 
family  background  information  from  some  national  surveys). 

The  following  section  highlights  the  main  findings  of  the  LFS  and  SLS  analyses. 
These  are  presented  separately  for  the  purposes  of  clarity  but  section  1.5  will  draw 
together  the  two  sets  of  findings  in  discussing  the  implications  of  the  CATEWE 
project  for  policy  and  future  research. 

1.4  MAIN  FINDINGS 

The  CATEWE  project  resulted  in  twenty-two  substantive  papers  on  different  issues 
relating  to  school  to  work  transitions,  including  the  nature  of  the  relationship  between 
education  and  labour  market  outcomes,  the  nature  of  apprenticeship  systems,  variation 
in  youth  programme  provision,  and  differentiation  among  groups  of  young  people  in 
terms  of  gender,  social  background  and  national  origin.  These  papers  draw  on 
complex  multivariate  analyses  which  enable  us  to  compare  European  countries  in  a 
systematic  way  and  allow  us  to  highlight  the  significant  dimensions  of  education  and 
training  influencing  the  transition  process.  The  remainder  of  this  section  presents  a 
broad  overview  of  the  main  project  findings. 

1.4.1  The  Eurostat  Labour  Force  Survey 

Analyses  of  the  Eurostat  Labour  Force  Survey  indicate  some  convergence  in  levels  of 
educational  attainment  across  Europe  with  the  greatest  increases  evident  in  countries 
with  previously  low  levels  of  attainment.  However,  there  is  no  evidence  of  any 
convergence  in  the  type  of  education  across  countries.  In  this  respect,  analyses 
identified  three  ideal-typical  groups  of  countries  within  Europe:  countries  with  an 
extensive  system  of  vocational  training  at  upper  secondary  level  (either 
apprenticeship-based  as  in  Germany  or  school-based  as  in  the  Netherlands)  and 


consequently  a high  prevalence  of  labour  market  entrants  with  occupationally-specific 
qualifications;  Northern  European  countries  with  fairly  large  proportions  of  entrants 
having  general  rather  than  vocational  qualifications;  and  Southern  European  countries 
with  lower  levels  of  educational  attainment  overall  and  less  vocational  specialisation. 

These  three  groups  of  countries  were  found  to  have  quite  distinct  patterns  of  labour 
market  integration  among  young  people.  Across  all  European  countries,  labour  market 
entrants  tend  to  experience  higher  unemployment  rates,  more  employment  instability 
and  have  lower  skilled  jobs  than  more  experienced  workers.  However,  the  gap 
between  these  groups  is  more  pronounced  in  Southern  European  countries  with 
entrants  being  vulnerable  to  proportionately  higher  unemployment  rates  and  more 
precarious  employment.  Northern  European  countries  were  found  to  occupy  an 
intermediate  position,  though  with  quite  a distinction  evident  between  'outsiders' 
(labour  market  entrants)  and  'insiders'  (more  experienced  workers)  particularly  in 
France,  for  example.  In  contrast,  unemployment  rates  among  labour  market  entrants 
closely  parallel  those  among  more  experienced  workers  in  systems  with  a strong 
emphasis  on  the  provision  of  occupationally-specific  skills,  such  as  Austria,  Germany, 
Denmark  and  the  Netherlands. 

Among  labour  market  entrants,  educational  level  is  found  to  be  highly  predictive  of 
transition  outcomes.  Those  with  lower  levels  of  education  have  significantly  higher 
unemployment  risks  than  the  better  qualified  across  most  European  countries,  with  the 
exception  of  Southern  Europe.  Furthermore,  educational  level  is  also  associated  with 
the  quality  of  employment  secured.  Those  with  higher  educational  levels  tend  to 
achieve  higher  occupational  status,  have  a lower  likelihood  of  entering  low-skilled  or 
temporary  jobs,  an  enhanced  access  to  professional  positions  and  are  more  likely  to 
obtain  full-time  contracts.  Type  of  education  is  also  crucial,  although  its  role  varies 
across  different  institutional  contexts.  Those  who  have  taken  part  in  vocational 
education/training  (especially  apprenticeships)  tend  to  have  a smoother  transition  to 
their  first  job  and  also  tend  to  access  more  stable  employment.  In  countries  without 
extensive  upper  secondary  vocational  training,  post-school  training  tends  to  reinforce, 
rather  than  compensate  for,  initial  levels  of  education. 


3 

ERIC 


7 


13 


Among  labour  market  entrants,  unemployment  risks  are  also  related  to  aggregate 
macroeconomic  conditions  with  the  least  qualified  being  particularly  vulnerable  to 
cyclical  swings.  Macroeconomic  conditions  have  a much  stronger  influence  on 
unemployment  risks  than  they  do  on  the  quality  of  the  job  obtained.  However,  there  is 
evidence  of  some  changes  in  job  quality  over  time;  educational  expansion  is 
associated  with  lower  net  returns  to  education  in  terms  of  occupational  status  and  skill 
level.  This  is,  at  least  partly,  counterbalanced  by  the  tendency  of  educational 
expansion  to  be  associated  with  an  increasing  professionalisation  of  the  labour  force. 

1.4.2  Analyses  of  SLS  data 

Analyses  of  school  leavers'  survey  data  focused  on  a narrower  range  of  countries  but 
the  availability  of  a wider  range  of  variables  and  multidimensional  indicators  of 
educational  and  transition  outcomes  enabled  us  to  explore  heterogeneity  among  this 
(mainly  Northern  European)  group.  As  with  analyses  of  the  Labour  Force  Survey, 
level  of  education  was  found  to  be  highly  predictive  of  transition  outcomes  among 
young  people.  Those  who  leave  school  with  lower  levels  of  education  have  a higher 
risk  of  being  unemployed  (immediately  after  leaving  school  and  over  the  first  five 
years  in  the  labour  market)  and  their  unemployment  spells  tend  to  be  longer  in 
duration.  Some  cross-national  variation  is  evident  in  the  distribution  of  unemployment 
with  unqualified  school-leavers  in  Ireland  experiencing  more  long-term 
unemployment  while  in  France  unemployment  tends  to  be  interspersed  with  periods 
of  participation  in  youth  programmes  or  short-term  employment.  When  they  secure 
employment,  the  least  qualified  tend  to  enter  part-time,  lower  status  and/or  lower 
skilled  jobs. 

Other  dimensions  of  educational  differentiation  are  found  to  influence  transition 
outcomes,  but  often  in  different  ways  in  different  institutional  systems.  Examination 
grades  are  found  to  have  a more  significant  effect  on  transition  outcomes  in  more 
general  education  systems  than  in  more  track-differentiated  systems.  Higher- 
performing students  have  reduced  unemployment  risks  in  Ireland,  Scotland  and 
Sweden  but  grades  are  not  significantly  associated  with  unemployment  chances  in  the 
Netherlands  where  type  and  level  of  education  play  a more  important  role. 


School-based  vocational  education  has  the  strongest  labour  market  effects  in  formally 
track-differentiated  systems  with  occupational ised  labour  markets  (such  as  the 
Netherlands).  However,  there  is  some  evidence  (as  in  the  Irish  case)  that  the 
development  of  occupationally-specific  vocational  courses  within  the  framework  of  a 
'general'  education  system  may  yield  similar  benefits.  Post-school  vocational 
education  in  the  form  of  apprenticeship  and  participation  in  youth  programmes  was 
also  considered.  The  apprenticeship  system  plays  a distinctive  role  across  countries, 
forming  an  alternative  to  school-based  vocational  education  as  a route  to  skills  in 
France  and  the  Netherlands  but  operating  as  a type  of  post-school  vocational  training 
in  Ireland  and  Scotland.  The  prevalence  of  youth  programmes  also  varies  across 
countries,  being  higher  in  Scotland,  France  and  Sweden  than  in  Ireland  or  the 
Netherlands.  In  part,  this  is  related  to  greater  labour  market  regulation  in  France  and 
Sweden  with  schemes  providing  work  experience  as  a means  of  labour  market  access. 
However,  in  Scotland  youth  programmes  have  become  an  important  route  to  skill 
acquisition  even  in  the  context  of  a relatively  'flexible'  labour  market. 

The  nature  of  the  education/training  system  influences  the  extent  of  differentiation  in 
terms  of  gender  and  ethnicity.  In  particular,  early  selection  into  different  tracks  plays 
a role  in  increasing  differentiation  in  terms  of  educational  and  transition  outcomes.  In 
the  Netherlands,  for  example,  gender  differences  in  the  type  of  vocational  tracks  taken 
is  associated  with  somewhat  higher  occupational  and  industrial  segregation  by  gender 
on  entry  to  the  labour  market.  However,  it  should  be  noted  that  gender  segregation 
within  the  labour  market  was  a feature  of  all  of  the  countries  considered.  In  a similar 
way,  early  selection  in  the  French  system  is  associated  with  greater  differences  in 
educational  outcomes  between  immigrant  and  native-born  young  people.  Analyses 
from  the  CATEWE  project  indicate  no  evidence  that  educational  and  transition 
outcomes  have  become  less  structured  over  time  by  background  characteristics  such 
as  gender,  social  class  and  ethnicity. 

1.5  IMPLICATIONS  FOR  POLICY  AND  RESEARCH 


ERIC 


Perhaps  the  main  conclusion  of  the  CATEWE  project  is  that,  given  the  diversity  in 
education,  training  and  labour  market  systems  across  Europe,  the  same  policy 
interventions  are  unlikely  to  be  equally  effective  in  different  contexts.  It  is  worth 


tc 


9 


noting,  however,  that,  while  relatively  enduring  in  the  way  they  shape  school  to  work 
transitions,  national  transition  systems  are  not  fixed.  Indeed,  many  of  the  countries 
considered  in  our  analyses  have  experienced  considerable  changes  in  recent  decades, 
particularly  in  the  nature  of  their  apprenticeship  and  youth  programme  provision. 
Transition  processes  and  outcomes  are,  therefore,  amenable  to  policy  intervention  but 
not  to  the  imposition  of  a single  'solution'  derived  from  a very  different  institutional 
context.  It  has  proved  difficult,  for  instance,  to  expand  the  apprenticeship  system 
beyond  traditional  crafts  sectors  in  certain  systems  (such  as  Ireland)  so  this  would  not 
appear  to  be  a viable  solution  to  early  educational  failure  in  all  contexts. 

A fairly  striking  regularity  across  Europe  is  the  crucial  role  of  educational  level  in 
shaping  transition  outcomes.  Those  with  low  levels  of  education/training  continue  to 
experience  marginalisation  within  the  labour  market  (either  through  increased 
unemployment  or  precarious  work  situations),  even  in  the  context  of  rapid 
employment  growth.  There  is,  therefore,  a need  for  policy  intervention  to  prevent 
drop-out  from  the  education  system  and/or  to  provide  alternative  routes  to  skills 
acquisition  and  labour  market  integration  for  young  people  who  have  experienced 
educational  failure. 

The  provision  of  vocational  education  may  be  seen  as  one  means  of  achieving  this 
end.  All  else  being  equal,  young  people  who  have  taken  a vocational  track  are  found 
to  have  a smoother  transition  to  the  labour  market  and  tend  to  access  better 
employment  opportunities.  This  pattern  is  particularly  marked  for  those  who  have 
taken  an  apprenticeship  programme.  But  the  potential  disadvantages  of  promoting 
greater  commitment  to  vocational  education  must  also  be  considered.  Young  people 
who  have  taken  vocational  tracks  are  less  likely  than  those  who  have  taken  general 
tracks  to  enter  higher  education  and,  in  the  longer  term,  they  tend  to  be  excluded  from 
higher  status  occupations.  Early  selection  into  different  educational  tracks  may  also 
exacerbate  social  differences  in  educational  outcomes,  leading  to  more  unequal 
outcomes  for  working-class  and  ethnic  minority  youth. 

Policy  interventions  thus  need  not  only  to  respond  to  'average'  transition  patterns  but 
to  take  account  of  diversity  among  young  people  in  terms  of  gender,  social  class  and 
ethnicity.  Discussions  about  gender  equity  have  often  focused  on  the  position  of 


women  'returners'  to  the  labour  force.  However,  our  analyses  indicate  the  persistence 
of  segregation  by  gender  among  young  women  in  the  face  of  the  introduction  of  equal 
opportunity  legislation  across  Europe.  More  research  is,  therefore,  needed  to  examine 
both  the  formal  and  the  informal  factors  shaping  choices  on  the  part  of  employers, 
education/training  providers  and  young  people.  There  is  also  a need  for  more 
information  on  the  social  class  and  ethnic  background  of  young  people  and  the 
development  of  equity  measures  in  education  and  transition  outcomes. 

Other  research  is  needed  to  inform  policy  in  a range  of  areas,  including  an 
investigation  of: 

• employer  strategies  in  different  segments  of  the  labour  market,  in  particular,  their 
decisions  in  relation  to  the  recruitment  and  training  of  young  people; 

• young  people's  own  views  of  the  transition  process,  their  expectations  and 
aspirations; 

• the  role  of  policy  interventions,  especially  youth  programmes,  in  the  transition 
process  and  their  relationship  with  other  forms  of  education  and  training; 

• the  role  of  field  of  education/training  in  shaping  transition  outcomes,  controlling 
for  level  of  education; 

• regional  and  local  differences  in  educational  and  transition  outcomes. 


It  is  recommended  that  the  potential  use  of  existing  data  sources  should  be  enhanced 
in  order  to  study  transition  processes  from  a comparative  perspective.  Firstly, 
improved  documentation  of,  and  access  to,  EULFS  microdata  would  facilitate 
transitions  research.  Secondly,  the  transitions  module  added  to  the  Labour  Force 
Survey  has  considerable  potential,  but  only  if  researchers  are  granted  access  to  the 
microdata.  In  addition,  the  potential  of  the  module  to  collect  information  on  social 
background  and  career  trajectories  should  be  enhanced  in  future  waves  of  the  survey. 
Thirdly,  while  full  harmonisation  of  existing  national  transition  surveys  is  not 
feasible,  it  is  recommended  that  agreement  should  be  reached  on  a template  which 
represents  best  practice  and  principles  for  the  partial  harmonisation  of  these  surveys. 
The  added  value  of  attempting  such  harmonisation  could  be  high,  particularly  as  the 
number  of  European  countries  conducting  such  surveys  has  been  increasing.  Finally, 
we  recommend  the  initiation  of  a European-wide  survey  based  on  a prospective  age 


0 

ERIC 


11 


17 


cohort  design  starting  at  about  fifteen  years,  followed  over  a period  of  about  ten  years. 
This  would  enable  us  to  examine  decision-making  processes  among  young  people  at 
the  point  of  leaving  compulsory  education  and  their  subsequent  trajectories  through 
the  education  and  labour  market  systems.  The  improvement  of  existing  data  sources 
coupled  with  the  collection  of  new  data  would  greatly  enhance  our  ability  to 
understand  transition  systems  across  Europe  in  years  to  come. 

1.6  DISSEMINATION 

A number  of  complementary  dissemination  strategies  were  adopted  within  the  life- 
time of  the  CATEWE  project.  These  included  the  presentation  of  conference  papers, 
publication  in  journal  and  book  format,  the  creation  of  a CATEWE  web-site 
containing  working  papers  (http://www.mzes.uni-mannheim.de/proiekte/catewe/'),  and 
the  development  of  links  with  OECD  and  other  relevant  organisations  in  relation  to 
proposals  for  the  harmonisation  of  transitions  data.  Future  activities  will  include  the 
publication  of  a book  on  the  Labour  Force  Survey  analyses,  the  publication  of  papers 
on  the  School  Leaver  Survey  analyses  in  scientific  journals,  along  with  the  use  of 
existing  networks  among  project  partners  to  carry  out  dissemination  to  policy-makers 
and  other  interested  parties  at  a national  level. 


CHAPTER  2:  BACKGROUND  AND  OBJECTIVES 
2.1  RATIONALE  FOR  THE  PROJECT 

In  recent  decades,  European  countries  have  experienced  significant  growth  in  the 
educational  qualifications  of  labour  market  entrants.  At  the  same  time,  occupational 
structures  and  employment  practices  have  undergone  considerable  change  across 
Europe.  Such  changes  highlight  the  necessity  of  understanding  the  relationship 
between  the  education/training  system  and  labour  market  changes  in  different  national 
contexts.  In  particular,  the  persistence  of  youth  unemployment  in  many  European 
countries  means  that  attention  should  be  given  to  the  period  of  early  labour  market 
integration.  It  is  also  important  that  the  transition  from  education  to  the  labour  market 
should  be  examined  in  comparative  perspective.  The  diversity  of  institutional 
arrangements  across  EU  member  states  means  that  we  cannot  assume  that  policy 
interventions  will  operate  in  the  same  way  in  different  contexts.  Instead,  we  need  to 
identify  the  important  dimensions  of  national  education,  training  and  labour  market 
systems  in  order  to  formulate  effective  policy  in  this  regard. 

A number  of  researchers  have  examined  institutional  variation  in  education,  training 
and  labour  market  systems,  proposing  that  the  nature  of  the  relationship  between  these 
elements  is  societally  specific  (see,  for  example,  Maurice  et  al.,  1986).  While  such 
work  has  improved  our  understanding  of  the  transition  process,  empirical  research  has 
proved  rather  limited  and  narrowly  focused.  Cross-national  studies  have  tended  to 
focus  only  on  two  or  three  countries,  running  the  risk  of  generalising  to  very  different 
institutional  contexts  on  the  basis  of  a limited  number  of  'core'  European  countries 
(usually  Germany,  France  and/or  Britain).  Measures  of  education  and  labour  market 
outcomes  have  often  obscured  very  real  differences  between  systems  and  many 
crucial  dimensions  of  the  transition  process  have  been  ignored.  The  CATEWE  project 
set  out  to  address  this  deficit  by  developing  a more  adequate  and  comprehensive 
conceptual  framework  for  examining  the  relationship  between  education,  training  and 
labour  market  systems,  and  applying  this  framework  to  a comparative  empirical  study 
of  recent  developments  in  several  European  countries  (see  Hannan  et  al.,  1999). 


13  19 


2.2  BACKGROUND  TO  THE  PROJECT 


The  project  developed  out  of  the  experiences  of  project  partners  in  participating  in  the 
European  Network  on  Transitions  in  Youth.  This  Network  was  established  in  1992, 
drawing  together  researchers  who  had  considerable  experience  in  conducting  and 
analysing  national  transition  surveys  as  well  as  researchers  who  had  used  Labour 
Force  Survey  data  to  analyse  the  relationship  between  educational  and  labour  market 
outcomes.  Participation  in  the  Network  increased  awareness  of  national  differences  in 
transition  processes  as  well  as  of  the  kind  of  information  available  to  study  transitions 
and  stimulated  a number  of  small  (usually  two-country)  comparative  studies'.  It 
quickly  became  clear  that  it  was  necessary  to  expand  the  range  of  countries 
considered  in  order  to  capture  the  most  important  dimensions  of  institutional 
variation. 

Discussions  around  this  topic  led  to  the  preparation  of  a paper  for  the  OECD,  which 
was  then  preparing  a Thematic  Review  of  the  Transition  from  Initial  Education  to 
Working  Life,  outlining  a preliminary  framework  for  analysing  institutional 
differences  in  school  to  work  transitions  (Hannan,  Raffe,  Smyth,  1997),  and  to  the 
formulation  of  a research  project  funded  under  the  Leonardo  project.  This  project 
drew  on  national  transition  surveys  to  focus  on  the  experience  of  'lower  level  leavers' 
(those  leaving  school  without  upper  secondary  certification)  in  four  countries:  France, 
Ireland,  the  Netherlands  and  Scotland  (see  Hannan,  Smyth  et  al.,  1998).  The 
experience  of  attempting  to  develop  an  integrated  cross-national  database  highlighted 
the  potential,  as  well  as  the  difficulties,  involved  in  such  an  exercise.  It  also  became 
clear  that  an  adequate  analysis  of  institutional  diversity  would  need  to  take  account  of 
a broader  range  of  national  systems.  As  a result,  the  CATEWE  project  aimed  to  draw 
on  two  complementary  sources  of  data:  the  Eurostat  Labour  Force  Survey  and 
national  school  leavers'  surveys. 


1 Proceedings  of  the  European  Network  on  Transitions  in  Youth  annual  workshops  have  been 
published  in  CEDEFOP  (1994),  ESRI,  Combat  Poverty  Agency  (1998),  Raffe  et  al.  (1999),  and 
Hammer  (2000). 


2.3  PROJECT  OBJECTIVES 


The  original  objectives  of  the  project  were: 

1.  To  develop  a more  adequate  and  comprehensive  conceptual  framework,  drawing 
on  existing  research  and  new  analyses,  of: 

• the  nature  of  national  systems  of  initial  education  and  vocational  training  in 
European  societies,  with  particular  reference  to  their  degree  of  differentiation 
and  standardisation  as  well  as  to  the  relationship  between  general  and 
vocational  education  in  each  system; 

• the  factors  and  processes  affecting  variation  in  the  full  range  of  education/ 
training  outcomes  by  different  groups  of  young  people  in  each  system; 

• the  processes  of  transition  from  initial  education/training  to  work,  the  main 
outcomes  of  such  transitions  and  pathways,  and  the  main  factors  affecting 
success  and  failure  in  such  transitions  in  each  system; 

• the  impact  of  national  institutional  differences  on  education/  training  outcomes 
and  transition  processes  among  young  people. 

2.  Using  this  conceptual  framework,  to  construct  an  integrated  cross-national  data  set 
using  national  school  leavers'  surveys  for  France,  Ireland,  the  Netherlands  and 
Scotland  (UK),  countries  with  widely  varying  education/training  systems  and 
labour  market  structures.  Strenuous  efforts  were  to  be  made  to  incorporate  other 
countries  who  had  conducted,  or  were  planning  to  conduct,  school  leavers’ 
surveys. 

3.  To  analyse  education  to  work  transitions  across  all  European  countries  using  the 
Labour  Force  Surveys,  placing  the  four  country  analysis  of  school  leaver 
transitions  (2)  in  the  context  of  the  wider  European  context. 

4.  Using  these  comprehensive  datasets, 

• To  test  and  refine  the  conceptual  framework  (1)  and  associated  hypotheses;  and 
to  develop  a more  adequate  and  comprehensive  framework  to  study  school  to 
work  transitions  across  all  European  countries; 


• To  explore  national  similarities  and  differences  in  ET  systems  and  their 
outcomes,  at  an  aggregate  level  and  for  different  groups  of  young  people;  and 
the  way  in  which  national  differences  in  these  respects  are  influenced  by 
institutional  factors; 

• To  identify  the  main  factors  influencing  success  or  failure  in  ET  outcomes  and 
labour  market  integration  in  each  system;  and  attempt  to  explain  similarities  and 
differences  in  these  patterns  across  the  different  national  systems. 

5.  To  develop  proposals  to  harmonise  existing  school  leavers'  surveys  in  the 
participating  countries;  and  encourage  the  extension  of  comparative  transition 
surveys  to  other  European  countries  currently  planning  surveys  of  school  leavers. 

It  was  intended  that  analyses  should  be  framed  in  terms  of  five  key  research 

questions: 

• What  is  the  nature  and  extent  of  similarities  and  differences  in  education/training 
systems  within  the  EU  countries  studied,  and  in  the  associated  type  and  level  of 
education  and  training  achieved  by  educational  system  leavers  entering  the  labour 
market? 

• What  is  the  relationship  between  differences  in  education/training  outcomes  and 
the  social  background  characteristics  of  system  leavers  (in  terms  of  gender,  social 
class  and  ethnic  origin)?  And  do  such  differences  vary  systematically  across 
national  systems? 

• How  do  transition  processes  vary  systematically  across  countries  (in  terms  of  their 

length,  complexity  etc.)?  And  to  what  extent  are  these  differences  related  to 
differences  in  education,  training  and  labour  market  structures? 

• What  is  the  nature  and  extent  of  the  relationship  between  level  and  type  of 
educational  achievements  among  system  leavers  and  (the  success  of)  their 
transition  processes  and  outcomes?  How  do  these  relationships  vary  by  type  of 
system? 

• What  is  the  relationship  between  social  background  characteristics  and  labour 
market  outcomes?  To  what  extent  is  this  relationship  mediated  by  education,  and 
does  this  vary  by  type  of  system? 


O 


16 


22 


The  following  section  outlines  how  these  objectives  were  met,  and  research  questions 
addressed,  during  the  course  of  the  project.  However,  the  objectives  themselves 
became  subject  to  review  and  development  over  the  course  of  the  project.  Firstly,  it 
became  clear  that,  at  least  in  terms  of  national  transition  survey  data,  information  was 
available  only  on  a limited  range  of  European  countries.  As  a result,  an  attempt  was 
made  to  broaden  the  scope  of  countries  included.  Information  from  a Swedish  cohort 
study  was  incorporated  into  the  school  leavers'  database(s)  and  links  were  forged  with 
two  countries  (Belgium  (Flanders)  and  Portugal)  who  were  initiating  transition 
surveys.  Furthermore,  analyses  of  the  Labour  Force  Survey  data  indicated  the 
distinctive  character  of  institutional  and  market  systems  in  Southern  Europe.  For  this 
reason,  analyses  of  microdata  for  Italy  and  Spain  were  conducted  in  order  to  further 
explore  the  source  of  such  variation.  Secondly,  the  construction  of  an  integrated 
database  using  school  leavers'  survey  data  revealed  the  rich  potential  of  this  data 
source.  As  a result,  four  separate  integrated  datasets  were  constructed,  allowing  us  to 
examine  cross-national  differences  at  one  point  in  time,  over  the  1980s  and  1990s, 
and  five  years  after  leaving  school.  Further  details  on  the  development  of  the  project 
are  presented  in  the  remainder  of  the  report. 


ERIC 


17 


23 


CHAPTER  3: 

SCIENTIFIC  DESCRIPTION  OF  PROJECT  RESULTS  AND  METHODOLOGY 

3.1  INTRODUCTION 

In  this  chapter  we  describe  the  main  activities  of  the  project  and  summarise  its  findings.  We 
start,  in  section  3.2,  by  outlining  the  project’s  conceptual  framework,  paying  most  attention  to 
dimensions  of  variation  in  national  ‘transition  systems’.  In  section  3.3  we  then  contrast 
different  approaches  to  comparative  research,  and  describe  the  core  strategy  of  the  project 
which  involves  what  we  describe  as  relatively  ‘intensive’  comparisons  between  countries. 
The  project  took  advantage  of  the  complementary  opportunities  provided  by  its  two  main  data 
sources,  the  Eurostat  Labour  Force  Survey  (EULFS)  and  national  School  Leavers'  Surveys 
(SLS),  and  these  are  described  in  the  following  three  sections.  In  section  3.4  we  discuss  the 
relative  strengths  and  weaknesses  of  these  two  data  sources.  In  section  3.5  we  summarise  our 
work  on  the  EULFS,  presenting  our  findings  in  terms  of  four  main  themes.  In  section  3.6  we 
similarly  summarise  our  work  on  school-leaver  surveys.  Finally,  in  section  3.7  we  describe 
the  project’s  work  on  new  data-collection:  its  support  for  new  transition  surveys  in  two 
European  countries  and  the  development  of  proposals  for  future  comparative  data  on 
transition,  but  on  this  and  other  issues  we  save  our  main  recommendations  for  chapter  4. 

3.2  CONCEPTUAL  FRAMEWORK 

The  research  questions  outlined  in  the  previous  chapter  can  be  summarised  as  a single  general 
question:  “How  do  national  transition  systems  shape  transition  processes  and  outcomes?”  To 
answer  this  question  we  needed  concepts  of  transition  system,  and  of  the  dimensions  of 
variation  in  national  systems;  and  we  needed  an  understanding  of  the  main  transition 
processes  and  outcomes. 

The  project’s  first  report  was  submitted  in  summer  1998  and  published  as  a Working  Paper  in 
the  following  year  (Hannan  et  al.,  1999).  It  presented  the  project’s  initial  conceptual 
framework,  which  was  to  inform  the  construction  of  datasets  and  the  analysis  of  data.  The 
framework  built  on  previous  research,  including  the  review  by  Hannan,  Raffe  and  Smyth 
(1997)  described  in  chapter  2,  and  on  country  studies  which  summarised  existing  research  on 


ERIC 


18 


education-to-work  transitions  in  France,  Germany,  Ireland,  The  Netherlands,  Portugal  and 
Scotland.  The  country  studies  were  published  at  the  same  time,  together  with  an  analysis  of 
the  changing  economic  and  demographic  context  of  transition. 

The  report  stressed  that  the  conceptual  framework  was  an  initial  statement,  for  subsequent 
review  and  refinement,  and  it  continued  to  develop  during  the  life  of  the  project.  In  the 
presentation  that  follows  we  focus  on  the  conceptual  framework  as  we  used  it  in  our  work. 
We  take  the  initial  framework,  presented  in  our  first  report,  as  the  starting  point  and  outline 
the  main  lines  of  development,  giving  most  emphasis  to  features  which  proved  important  in 
our  empirical  analyses. 


The  initial  conceptual  framework  identified  three  elements  (or  groups  of  dimensions)  in 
cross-national  comparisons  of  transition: 

• the  demographic,  economic  and  labour-market  context  within  which  transitions  occur; 

• the  dimensions  of  the  education/training  system; 

• the  nature  of  transition  processes  and  outcomes. 

As  the  project  developed  we  found  it  useful  to  combine  parts  of  the  first  two  elements  to 
develop  the  concept  of  ‘transition  system’.  This  embraced  the  main  independent  variables  of 
our  analyses:  the  macro-level  ‘determinants’  of  transition  patterns.  The  third  element  of  our 
framework  provided  the  main  dependent  variables,  the  processes  and  outcomes  of  transitions 
at  the  micro  level.  The  concept  of  ‘transition  system’  describes  the  relatively  enduring 
features  of  a country’s  institutional  and  structural  arrangements  which  shaped  transition 
processes  and  outcomes.  Our  work  paralleled  that  of  the  OECD’s  Thematic  Review  of  the 
Transition  from  Initial  Education  to  Working  Life  in  developing  understandings  of  transition 
systems  (OECD,  2000).  The  paper  by  Hannan  et  al.  (1997)  was  written  for  the  OECD 
Review,  and  members  of  the  project  and  the  OECD  had  a joint  meeting  in  1998. 


3.2.1  Dimensions  of  transition  systems  and  their  contexts 


With  respect  to  the  demographic,  economic  and  labour-market  context  (the  first  of  the  three 
elements  listed  above),  the  report  identified  the  following  key  dimensions: 

• the  ratio  of  young  people  to  adults,  both  in  the  population  as  a whole  and  in  the  labour 
force,  including  the  effect  of  migration; 


3 

ERIC 


19 


OR 


• the  nature  and  resources  of  the  family  system,  and  how  this  provides  opportunities  or 
encouragement  to  young  people  to  stay  at  home  or  to  set  up  independent  households; 

• the  nature  of  economic  development:  this  embraces  the  distinction  between  the  economic 
core  and  periphery  of  Europe,  and  variations  in  the  structure  and  ownership  of  firms; 

• the  stage  in  the  economic  cycle; 

• the  industrial  and  occupational  structure  of  employment; 

• the  nature  of  labour-market  segmentation  within  a country,  with  particular  reference  to  the 
relative  importance  of  occupational  labour  markets  (OLMs),  internal  labour  markets 
(ILMs)  and  external  labour  markets  (ELMs).  These  distinctions  are  complex  and  countries 
constitute  varying  mixtures  of  all  three  types; 

• the  size  and  nature  of  the  informal  economy; 

• the  extent  of  regulation  of  the  labour  market; 

• the  age  structuring  of  the  labour  force:  that  is,  the  extent  of  segregation  between  youth  and 
adult  labour  markets;  and 

• the  gender  and  ethnic  structuring  of  the  labour  force,  and  the  strength  of  segregation 
associated  with  these  traits. 

The  project  focused  especially  on  the  dimensions  of  labour-market  structure,  including  the 
nature  of  segmentation  (the  OLM/ILM  distinction)  and  the  extent  of  regulation.  Together 
with  features  of  education  and  training  systems,  described  below,  these  were  the  main 
elements  of  the  concept  of  transition  system  as  it  was  employed  in  the  project’s  analyses.  The 
ILM/OLM  distinction  is  central  to  a variety  of  distinct  but  related  theoretical  approaches, 
including  Maurice,  Sellier  and  Silvestre’s  (1986)  distinction  between  qualification  space  and 
organisational  space,  Marsden’s  (1986)  analysis  of  labour-market  segmentation  and  Garonna 
and  Ryan’s  (1991)  analysis  of  modes  of  inclusion  and  exclusion  in  youth  labour  markets. 

Many  of  the  dimensions  listed  above  are  of  interest  more  on  account  of  their  variation  over 
time,  than  of  their  variation  between  countries.  For  example,  transition  processes  and 
outcomes  are  influenced  by  the  stage  of  the  economic  cycle  and  by  trends  in  the  occupational 
or  industrial  structure  of  employment. 

With  respect  to  the  second  element,  the  dimensions  of  education  and  training  systems,  the 
project’s  first  report  identified  the  following  dimensions  or  sets  of  dimensions: 


• standardisation:  the  extent  to  which  curricula,  assessment  and  certification,  and  related 
quality  assurance  procedures,  are  standardised  on  a national  or  regional  basis. 
Standardisation  may  vary  across  different  parts  of  the  same  system:  for  example 
vocational  qualifications  may  be  more  standardised  than  general  qualifications,  or  vice 
versa.  However  the  project  made  less  use  of  the  concept  of  standardisation  in  its  empirical 
analyses,  partly  because  of  the  limited  variation  across  European  countries; 

• differentiation:  the  extent  and  manner  in  which  a system  differentiates  its  students.  The 
initial  conceptual  framework  noted  that  this  might  differ  between  stages  of  education 
within  the  same  system.  It  further  distinguished  three  types  of  differentiation,  which 
respectively  concerned 

♦ institutions  or  programmes  within  a stage, 

♦ progression  into  the  next  stage,  and 

♦ outcomes  at  the  end  of  a stage,  especially  the  level  of  attainment  or  grades 
achieved. 

The  project’s  empirical  analyses  confirm  the  multi-dimensional  nature  of  differentiation, 
but  do  not  settle  on  a single  way  to  represent  these  dimensions.  Many  analyses  work  with 
a simpler  concept  of  track  differentiation,  which  broadly  corresponds  to  the  first  two  types 
of  differentiation  listed  above,  as  applied  to  Tower’  and  ‘upper’  secondary  education  and 
the  relation  between  them.  However  some  systems  with  relatively  low  levels  of  track 
differentiation,  such  as  Ireland  or  Scotland,  may  have  strong  vertical  differentiation  with 
respect  to  attainment  at  the  end  of  each  stage,  so  the  project  sometimes  found  it  useful  to 
distinguish  outcome  differentiation  (attainment  at  the  end  of  a stage)  from  track 
differentiation; 

• school-to-work  linkages:  the  role  of  employers  in  the  education/training  system.  At  one 
pole  are  systems  where  employers  are  direct  providers  of  education/training,  for  example 
through  apprenticeship.  Other  systems  (such  as  the  Netherlands)  have  collinear 
relationships,  with  employers  making  an  important  institutionalised  input  into  school- 
based  vocational  education.  A third  category  (such  as  Japan)  comprises  systems  where 
employers  have  links  with  schools  for  recruitment.  At  the  other  pole  are  systems  where 
employers  have  little  direct  involvement  in  any  of  the  senses  described  above.  These 
comprise  two  sub-categories:  systems  in  which  recruitment  decisions  are  substantially 
based  on  individuals’  educational  attainments,  which  therefore  send  out  strong  market 
signals  to  education;  and  systems  where  this  does  not  happen  and  market  signals  are 

o 

ERIC 


21 


weak.  Different  types  of  linkage  may  be  found  in  different  sectors  of  the  same  education 
and  training  system,  but  there  is  cross-national  variation  in  the  linkages  that  are 
characteristic  of  equivalent  sectors; 

• youth  training:  arrangements  for  youth  training,  and  work-based  provision  more 
generally,  vary  with  respect  to  the  level  of  provision,  the  degree  to  which  it  is 
differentiated  (including  between  apprenticeship  and  other  programmes)  and  the  formal 
inclusiveness  of  provision; 

• in  addition,  several  CATEWE  analyses  have  examined  the  level  of  participation  and/or 
attainment  in  education,  with  particular  reference  to  the  relative  scale  of  lower-secondary, 
upper-secondary  and  tertiary  education,  or  their  ISCED  equivalents. 

These  dimensions  are  neither  logically  nor  empirically  independent  of  each  other.  We  discuss 
their  interconnections  in  the  next  section. 

3.2.2  Classifying  transition  systems 

The  project’s  first  report  drew  on  the  country  studies  to  produce  a provisional  classification  of 
European  countries  in  terms  of  the  dimensions  of  the  conceptual  framework  (Hannan  et  al., 
1999,  Figure  A.  1).  This  classification  is  detailed  and  we  do  not  reproduce  it  here,  but  it  draws 
attention  to  the  importance  of  variation  within  each  system  with  respect  to  many  of  the 
dimensions  examined.  For  example,  features  such  as  standardisation,  differentiation  or 
linkages  with  the  labour  market  may  vary  between  different  educational  tracks,  as  well  as 
between  different  stages  of  education,  within  the  same  system.  However  this  classification 
was  too  detailed  to  form  the  main  basis  of  the  project’s  empirical  comparisons,  and  it  was 
necessary  to  develop  a simpler  classification  with  which  to  work. 

When  countries  are  classified  according  to  a matrix  which  combines  many  of  these 
dimensions  (see  Figure  3.1)  they  form  a broad  continuum,  from  countries  with  high  levels  of 
standardisation,  strong  track  differentiation,  strong  linkages  and  significant  apprenticeship 
systems,  to  countries  with  weaker  track  differentiation  (but  often  strong  vertical  ‘outcome’ 
differentiation),  weak  linkages  and  in  some  cases  extensive  youth  programmes.  This 
continuum  is  also  associated  with  some  of  the  labour-market  dimensions  discussed  above,  and 
especially  with  the  distinction  between  OLMs  and  ILMs.  OLM  countries  tend  to  have 

ERJC 


22 


28 


standardised,  track-differentiated  education  systems,  strong  linkages  and  high  levels  of 
apprenticeship,  and  they  therefore  appear  at  the  former  end  of  the  continuum  of  transition 
systems  described  above. 

Of  the  countries  covered  by  the  CATEWE  country  studies,  Germany  followed  by  the 
Netherlands  is  at  one  end  of  the  continuum  with  strong  OLMs,  track  differentiation  and 
linkages,  while  the  others  (France,  Ireland,  Portugal  and  Scotland)  are  towards  the  latter  end 
but  in  the  category  with  strong  market  signals  from  the  labour  market  to  education.  All  are 
relatively  standardised.  However  the  project’s  first  report  acknowledged  that  the  distinctive 
features  of  the  transition  system  of  Portugal,  as  well  as  other  southern  European  countries, 
were  not  fully  captured  by  this  framework,  and  identified  this  as  an  area  for  further  study.  In 
addition,  a study  would  need  to  include  non-European  countries,  such  as  the  United  States 
and  Japan,  in  order  to  cover  the  full  range  of  variation  in  transition  systems  expressed  by  the 
conceptual  framework. 

Another  important  dimension  of  variation  in  national  labour  markets,  the  strength  of  labour- 
market  regulation  or  flexibility,  partly  cuts  across  this  continuum.  The  report  suggested  that 
this  could  be  measured  by  the  extent  of  state  regulation  of  employment  standards  and 
employment  protection,  and  by  the  extent  of  corporate  coverage  of  trade  union/employer 
relationships,  although  later  studies  refined  these  criteria.  The  strength  of  labour-market 
regulation  differentiates  among  the  large  block  of  ‘loosely  coupled’  countries  with  strong 
labour  market  signals.  Of  the  main  countries  studied  by  the  project,  Scotland  and  (to  a lesser 
extent)  Ireland  have  weaker  regulation  than  the  others. 

Different  CATEWE  analyses  emphasise  different  dimensions  of  variation  in  transition 
systems.  This  depends  partly  on  their  theoretical  starting  point,  and  partly  on  the  data  source 
and  the  opportunities  that  it  provides.  Analyses  of  EULFS  data  tend  to  focus  on  dimensions  of 
labour-market  variation,  in  particular  the  ILM/OLM  distinction,  rather  than  on  variation  in 
education  and  training  systems.  The  EULFS  provides  good  labour-market  outcome  measures 
with  which  to  test  the  ILM/OLM  distinction  and  to  derive  empirically-based  clusters  of 
countries.  The  EULFS  analyses  also  distinguished  countries  in  terms  of  the  strength  of 
labour-market  regulation,  and  time-varying  features  of  transition  systems  or  their  contexts 
such  as  levels  of  educational  participation,  occupational  composition  and  economic  activity. 


Conversely,  analyses  of  SLS  data  tend  to  focus  on  the  dimensions  of  variation  in  education 
and  training  systems.  The  main  dimensions  of  transition  systems  investigated  by  SLS 
analyses  are  track  differentiation,  linkages  and  the  role  of  work-based  provision. 

A theoretical  challenge  in  the  analysis  of  transition  systems  is  to  clarify  the  connection 
between  their  educational  and  labour  market  dimensions.  For  example,  is  the  connection 
between  dual  systems  and  OLMs  a necessary  connection,  explicable  in  terms  of  the 
theoretical  frameworks  used  to  analyse  labour  markets  and  education  systems,  or  is  it  merely 
a product  of  contingent  historical  circumstances?  However  these  connections  could  not  be 
fully  explored  during  the  project,  partly  because  of  the  different  emphases  of  the  two  main 
data  sources,  and  further  analysis,  perhaps  using  different  data,  is  required. 


24 


30 


Figure  3.1:  A typology  of  ET  systems  and  labour  market  linkages 


School-Work  Linkage 


(a)Tightly  coupled 
ET/employer  systems: 
strong  linkage 
(dual  system) 


Degree  of  Standardisation  of  ET  System 


High 


Low 


Degree  of  Differentiation  (and  Vocat./Occupat.  Specificity)  of  ET  Systems 


High^ 

Germany 

Austria 

Switzerland 

Denmark 


► Low 


High* 


Low 


Substantial  sharing  and  co- 
operation between  providers 
and  employers  in  delivery  of 
ET.  As  in  apprenticeships. 
High  occupationalisation  of 


LM 

(b)  Tightly  coupled 
ET/employer  systems: 
collinear  linkage: 


Netherlands 


High  levels  of  in-school  provision 
of  ET  specific  to  particular 
occupations,  agreed  with  employers. 
High  occupationalisation  of  LM. 

(c) Loosely  coupled  or 
decoupled  ET/employer 
systems,  but  with  strong 
market  signals: 

Low  degree  of  ET  provider  and 
employer  sharing  of  ET  provision; 
low  occupationalisation  of  LM,  and 
limited  school  involvement  in 
employment  decisions. 

(d) Loosely  coupled 
systems,  but  with  strong 
market  signals  and 
strong  school  placement 


England/Wales 

Scotland 

Italy 

France 

Portugal 

Finland 

Sweden 

Ireland 


function 

(e)De-coupled  ET/LM 
systems  with  weak 
market  signals  (from 
second  level). 


Japan 


USA 

Canada 


3.2.3  Transition  processes  and  outcomes 


With  respect  to  the  main  dependent  variables  of  the  analysis,  transition  processes  and 
outcomes,  the  project’s  first  report  notes  that  the  process  of  transition  has  changed  in  most 
countries,  involving  a longer  and  more  complex  sequence  of  transitions  and  a larger  number 
of  intermediate  statuses  between  education  and  a ‘stable’  status  in  the  labour  market.  There  is 
a debate  about  the  extent  to  which  youth  transitions  have  become  more  individualised  and 


ERjt 


25 


31 


whether  this  has  affected  the  nature  or  importance  of  social  and  gender  inequalities.  Important 
features  of  the  transition  process  include: 

• the  number  of  separate  transitions  which  comprise  a ‘completed’  transition  from 
education  to  work  for  an  individual; 

• the  length  of  the  period  over  which  these  transitions  take  place; 

• the  extent  of  differentiation  between  transition  statuses  (for  example,  the  extent  to  which 
apprenticeships  are  distinct  from  other  work-based  training,  and  the  extent  and  the  nature 
of ‘dual  statuses’  such  as  combinations  of  work  and  education); 

• the  nature  of  trajectories:  particularly  the  ways  in  which  education,  training,  qualification 
outcomes  and  employment/unemployment  are  interrelated  (for  example,  the  phasing  of 
education/training  and  employment  status  changes,  the  possibility  of  ‘reversals’  and 
‘bridging  loops’  back  to  education/training  from  the  labour  market,  and  the  extent  of 
education/training  involvement  leading  to  qualifications); 

• inequalities  in  respect  of  gender,  social  class  and  ethnicity;  and  relatedly, 

• the  extent  of  individualisation,  in  the  sense  either  of  a growth  in  the  number  and 
complexity  of  transitions,  or  a reduction  in  the  correlation  between  transition  processes 
and  background  characteristics  such  as  gender  and  social  class. 

The  concept  of  transition  outcome,  and  especially  the  definition  of  a successful  transition,  are 
problematic.  In  many  studies  the  available  data  make  it  necessary  to  define  outcomes  in  terms 
of  ‘snapshots’  at  a given  point  in  the  transition  process,  for  example  one  year  or  five  years 
after  leaving  school.  The  main  ‘snapshot’  outcomes  of  concern  to  this  study  are:  principal 
economic  activity;  occupational  status;  industrial  allocation;  labour  market  segment;  wages; 
security  of  employment;  access  to  on-the-job  training;  access  to  off-the-job  training 
sponsored  by  employers;  content  congruence,  that  is,  matching  between  type  of  education  and 
type  of  occupation;  ‘level  congruence’,  or  the  extent  of  matching  between  level  of  education 
and  occupational  status.  However  comparisons  based  on  snapshot  measures  are  sensitive  to 
the  time  at  which  they  are  measured.  The  difference  in  outcomes  between  young  people  from 
different  countries  may  be  very  different  one  year  after  leaving  school  but  may  become  more 
similar  a few  years  later.  More  fundamentally,  any  comparison  based  on  a single  measure  is 
problematic  in  the  context  of  cross-national  differences  in  the  length  of  the  transition  process, 
in  the  characteristic  sequence  of  transitions  and  in  the  blurring  between  statuses.  Other 
outcome  measures  try  to  capture  aspects  of  the  transition  sequence,  for  example  time  to  first 
O 

ERJC 


26 


job,  percentage  of  time  unemployed,  job  and  career  mobility,  frequency  of  job  changes  or  loss 
of  employment;  others  measure  more  complex  ‘trajectories’  of  status  changes  over  time. 
However  there  is  a need  for  further  conceptual  development  in  this  area. 

Finally,  the  project’s  first  report  refers  to  various  dimensions  of  state  intervention  (in  military 
service  requirements,  youth  programmes,  social  welfare  provision  as  well  as  the  labour 
market)  in  its  discussion  of  transition  processes  and  outcomes,  although  these  more  properly 
belong  with  the  concept  of  transition  system  rather  than  with  transition  processes  and 
outcomes. 

The  way  in  which  the  project  operationalises  transition  processes  and  outcomes  depends  on 
the  opportunities  afforded  by  the  data,  discussed  later  in  this  chapter.  Most  analyses  both  of 
the  EULFS  and  of  the  SLS  use  ‘snapshot’  measures  of  such  outcomes  as  (un)employment  or 
occupational  level.  The  EULFS  has  insufficient  longitudinal  data,  and  the  SLS  data  are 
insufficiently  comparable,  to  permit  much  ambition  in  defining  trajectories  or  other  measures 
of  the  sequencing  of  transition  statuses.  Analyses  of  both  data  sources  use  their  potential  for 
correlating  transition  outcomes  with  educational  background,  or  with  gender  and  social 
background. 

3.2.4  Summary:  conceptual  framework 

To  summarise,  the  conceptual  framework  developed  in  the  project: 

• uses  the  concept  of  transition  system  to  describe  the  interrelated  features  of  education 
and  training  systems,  national  labour  markets  and  other  macro-level  determinants  of 
transition  processes  and  outcomes; 

• identifies  a large  number  of  dimensions  of  transition  systems  and  of  their  social, 
demographic  and  economic  contexts; 

• suggests  that  many  of  these  can  be  expressed  in  terms  of  one  overarching  dimension, 
associated  with  the  strength  of  OLMs,  apprenticeship,  standardisation,  track 
differentiation  and  education/labour-market  linkages; 

• identifies  other  important  dimensions  of  variation  in  transition  systems,  including  the 
strength  of  labour-market  regulation  and  the  ‘vertical’  differentiation  of  levels  of 
educational  achievement  or  grades; 

O 

ERJC 


27 


33 


• acknowledges  the  need  for  further  theoretical  and  empirical  work  on  the  distinctive 
features  of  southern  European  transition  systems; 

• acknowledges  that  the  precise  connection  between  the  educational  and  labour-market 
dimensions  of  transition  systems,  and  the  distinction  between  a transition  system  and 
its  socio-economic  context,  require  further  analysis;  and 

• identifies  transition  processes  and  outcomes  in  several  ways:  in  terms  of  ‘snapshots’ 
of  individuals’  statuses  at  given  time  points,  classifications  of  longitudinal  transition 
sequences,  the  strength  of  association  between  education  and  labour-market  outcomes, 
and  social  and  gender  inequalities. 

3.3  COMPARATIVE  RESEARCH  STRATEGY 

The  project  required  comparable  data  on  transition  processes  and  outcomes  across  a sample  of 
countries  which  represented  theoretically  significant  variation  in  transition  systems.  We  used 
two  main  sources  of  data:  the  Eurostat  Labour  Force  Survey,  and  datasets  constructed  from 
national  surveys  of  secondary  school  leavers  in  France,  Ireland,  the  Netherlands,  Scotland  and 
Sweden.  The  project  aimed  to  benefit  from  the  complementary  strengths  of  these  two  sources 
of  data,  which  we  discuss  further  (together  with  their  respective  limitations)  in  section  3.4. 

There  are  different  ways  in  which  researchers  may  pursue  a question  such  as  “How  do 
national  transition  systems  shape  transition  processes  and  outcomes?”  Figure  3.2,  based  on 
Raffe  (2001),  summarises  contrasting  strategies  of  comparative  research.  At  root  they  reflect 
the  different  possible  purposes  of  comparison:  do  we  compare  countries  in  order  to  identify 
universal  laws  or  patterns,  or  in  order  to  elucidate  national  uniqueness?  (Kohn,  1987;  0yen, 
1990;  Bynner  and  Chisholm,  1998;  Evans,  1999).  A universalistic  approach  aims  to  ‘replace 
countries  by  variables’  (Ragin  and  Becker,  1992),  and  to  identify  a set  of  laws  which  not  only 
transcend  national  differences  but  also  explain  them.  It  would  attempt  to  answer  our  research 
question  by  reducing  differences  in  national  transition  systems  to  a series  of  dimensions  or 
types,  which  explain  why  transition  processes  and  outcomes  vary  across  countries  without  the 
need  to  refer  to  idiosyncratic  national  features.  By  contrast,  particularistic  research  aims  to 
discover  the  unique  logic  which  governs  social  processes  within  each  country.  Each  transition 
system  comprises  a unique  set  of  structures,  concepts  and  relationships  which  defy  any 
attempt  to  generalise  or  classify  across  countries;  even  phenomena  which  appear  to  be 
O 

ERIC 


28 


34 


general,  such  as  entry  to  the  labour  market  or  the  institution  of  apprenticeship,  have  different 
significance  in  terms  of  their  national  logics,  and  the  task  of  research  is  to  unpick  differences 
between  superficially  similar  concepts.  The  universalistic  strategy  typically  involves  an 
extensive  approach  to  comparison,  which  uses  large  samples  of  countries  in  order  to 
distinguish  empirically  among  alternative  country-level  explanatory  variables.  The 
particularistic  strategy  typically  uses  a small  sample  of  countries,  often  just  two  or  three,  to 
make  interpretive  comparisons,  whose  main  aim  is  to  highlight  qualitative  differences  in 
concepts  and  institutions.  It  is  this  approach  which  underlies  the  view  of  comparative  research 
as  a means  to  gain  a better  understanding  of  one’s  own  country,  by  exposing  taken-for- 
granted  assumptions  and  opening  them  to  challenge. 


Figure  3.2  Comparative  research  strategies 


Universalistic 

Intermediate/mixed 

Particularistic 

Aim 

To  identify  universal 
laws  or  patterns 

Mixed  aims 

To  elucidate  national 
uniqueness 

Method 

Replace  countries  with 
variables 

Use  common  concepts  to 
describe  and  classify 
national  logics  and 
analyse  differences 

Use  distinctive  concepts 
to  describe  and  analyse 
internal  logic  of  each 
country 

Use  of 
comparison 

Extensive: 

large  sample  of 
countries  to  represent 
variables  of  theoretical 
interest 

Intensive: 

use  small  sample  of 
countries;  multiple 
comparisons  provide 
degrees  of  freedom 

Interpretive: 

use  small  number  of 
national  contrasts  to 
highlight  differences  in 
concepts  and  institutions 

The  contrast  described  above  is  a matter  of  emphasis;  there  are  few  pure  examples  of  either 
strategy.  Many  researchers  have  adopted  an  intermediate  position  (eg  Maurice  et  al.,  1986; 
Kohn,  1987),  and  so  has  the  CATEWE  project.  In  the  middle  column  of  Figure  3.2  we 
identify  an  intermediate  strategy  which  may  have  both  universalistic  and  particularistic 
purposes,  which  recognises  the  existence  of  distinctive  national  ‘logics’  but  which  tries  to 
develop  common  cross-national  concepts  to  describe  them  and  cross-national  theories  which 
at  least  partially  explain  them.  Its  characteristic  research  approach  is  the  use  of  intensive 
comparisons,  which  test  a range  of  predicted  contrasts  or  similarities  across  a small  number  of 

EKIC 

r 


29  35 


theoretically  sampled  countries.  The  intensive  approach  compensates  for  the  lack  of  degrees 
of  freedom  at  the  country  level  by  making  multiple  comparisons  and  testing  a range  of 
hypotheses  arising  from  the  same  theoretical  starting  point.  The  intensive  approach  is 
therefore  dependent  on  detailed  comparable  data,  and  on  a strong  conceptual  and  theoretical 
foundation. 

Of  the  two  main  data  sources  used  by  the  CATE  WE  project,  only  the  Eurostat  LFS  provides  a 
basis  for  extensive  comparisons.  It  has  standardised  data  on  education  and  labour-market 
outcomes  for  all  fifteen  member  states  of  the  European  Union,  whereas  the  school-leaver 
survey  data  are  only  available  for  five  countries.  However  a sample  of  fifteen  countries  is  still 
a small  sample  to  be  used  for  extensive  comparisons,  especially  when  some  countries  have  to 
be  excluded  from  particular  analyses  for  reasons  of  sample  size  or  data  availability,  although 
the  degrees  of  freedom  can  be  increased  by  using  data  from  a sequence  of  annual  surveys.  As 
a result,  the  EULFS  is  most  effective  in  the  analysis  of  transition  systems  when  these  are 
conceptualised  in  terms  of  a small  number  of  types  (such  as  OLM  countries,  ILM  countries 
and  southern  European  countries:  see  3.5  below)  or  represented  by  time-varying  variables 
such  as  the  occupational  distribution  or  level  of  educational  participation.  In  the  CATEWE 
project,  analyses  of  the  EULFS  include: 

• mainly  descriptive  comparisons,  for  example  of  ‘dual  statuses’  which  combine  education 
with  work,  or  of  trends  in  educational  attainments  or  labour-market  outcomes; 

• analyses  which  distinguish  a small  number  of  types  of  transition  systems  among  EU 
countries; 

• multi-level  analyses  which  use  country-level  variables  (usually  including  time-varying 
variables)  along  with  individual  variables,  such  as  education,  to  predict  such  outcomes  as 
unemployment  or  the  type  of  employment  of  recent  labour-market  entrants. 


The  second  and  third  of  these  sometimes  correspond  to  our  notion  of  extensive  comparisons. 
However,  many  of  the  analyses  more  closely  resemble  intensive  comparisons,  as  they  take 
advantage  of  the  EULFS’  rich  data  across  a range  of  labour-market  outcomes  in  order  to 
make  the  multiple  comparisons  more  characteristic  of  the  intensive  approach.  Further 
analyses  use  data  from  national  labour  force  surveys  in  order  to  take  advantage  of  the  wider 
range  of  data  and  the  scope  for  re-defining  variables  for  more  focused  and  detailed 
comparisons. 


30 


36 


SLS  data  were  available  for  only  five  countries,  and  for  many  analyses  data  were  available 
only  for  a subset  of  these.  The  comparative  approach  in  the  SLS  analyses  has  been  intensive 
rather  than  extensive,  focusing  on  multiple  comparisons  across  a range  of  outcomes.  An 
example  is  Smyth’s  (2000a)  analysis  of  the  effects  of  varying  levels  of  differentiation  in 
education  systems  across  a range  of  indicators  of  gender  inequality  - in  the  level  of  education, 
type  of  education,  occupation,  income,  and  so  on.  The  SLS  data  for  many  transition  processes 
and  outcomes  (the  dependent  variable)  are  incomplete  or  insufficiently  comparable  across 
countries  (see  3.4  below).  The  school  leavers'  surveys  support  some  important  analyses  but 
their  potential  for  intensive  comparisons,  which  require  detailed,  comparable  data  on  a range 
of  processes  and  outcomes,  is  more  limited  than  we  had  hoped.  On  the  other  hand,  the  school 
leavers'  surveys  have  proved  unexpectedly  valuable  for  what  we  have  termed  interpretive 
comparisons.  The  processes  of  constructing  an  integrated  dataset,  and  of  defining  cross- 
nationally  applicable  measures  of  such  concepts  as  the  level  of  education  or  of  educational 
attainment,  raise  issues  of  meaning  and  equivalence  which  are  commonly  overlooked  in 
comparative  research.  And  deeper  investigations  of  concepts  and  institutions  such  as 
apprenticeship,  youth  programmes  and  upper-secondary  vocational  education  draw  attention 
to  cross-national  differences  in  their  organisation  and  their  role  in  the  transition  process. 

The  project’s  main  comparative  strategy  is  thus  an  intermediate  one  in  terms  of  Figure  3.2.  It 
uses  common  concepts  to  analyse  country  differences  and  (at  least  partly)  to  explain  them, 
and  it  relies  primarily  on  an  intensive  approach  which  tests  multiple  hypotheses  from  a given 
theoretical  starting  point  on  a small  sample  of  countries. 

3.4  THE  DATA  SOURCES  - THEIR  STRENGTHS  AND  WEAKNESSES 

This  section  discusses  the  strengths  and  weaknesses  of  the  two  data  sources  used  in  the 
project.  It  has  been  pointed  out  in  section  3.3  that  the  European  Union  Labour  Force  Survey 
(EULFS)  can  be  used  for  extensive  comparison  while  the  national  school  leavers'  (or 
transition)  surveys  are  useful  for  intensive  comparison.  The  data  sources  differ  in  other  ways 
which  are  outlined  in  Figure  3.3  (see  also  Raffe,  2000). 


31  3? 


Figure  3.3:  Comparison  of  data  sources 


Characteristics 

EU  Labour  Force  Survey 
(uptoLFS  1997) 

National  School  Leavers' 

Surveys 

Nature  of  survey 

Cross-sectional  (snapshot 
at  one  point  in  time) 

Longitudinal  (flow  out  of 
school) 

Data  structure 

Cross-sectional  with  only  a 
limited  retrospective 
component 

More  complete 
retrospective  information 
on  educational  and  labour 
force  histories;  some  panel 
(follow-up)  information 

Frequency 

At  least  annual 

Regularly,  though  not 
necessarily  annual 

Country  coverage 

All  EU  countries 

France,  Ireland,  the 
Netherlands,  Scotland  and 
Sweden1 

Sample  coverage 

All  adults;  allows  for  a 
comparison  of  young 
people  and  adults 

Young  people 
experiencing  the  transition 
only 

Comparability 

Constructed  to  a 
comparable  framework  but 
process  inadequately 
documented 

Not  designed  to  be 
comparable  but  useful 
comparative  indicators  can 
be  constructed 

Form  of  data 

Aggregate  only  at  EU 
level;  micro-data  for  some 
individual  countries 

Individual-level  data 

Information  on  educational 
background 

Level  and  type  (general  v. 
vocational) 

Level,  type,  field  and 
grades  (for  some  countries 
at  least) 

Information  on 
employment  position 

Detailed:  principal  status 
and  nature  of  job  but  lack 
of  information  on 
participation  in  youth 
programmes  or  earnings 

Detailed:  principal  status, 
nature  of  job,  earnings, 
and  participation  in  youth 
programmes 

Information  on  social 
background 

Not  available 

Parental  social  class, 
parental  education, 
immigration  status  (for 
some  countries  at  least) 

Perhaps  the  key  difference  is  that  the  EULFS  gathers  information  on  a cross-sectional  basis 
collecting  data  on  adults  within  households  at  a single  point  in  time.  The  EULFS,  therefore, 
has  little  direct  information  on  the  transition  process  itself,  but  researchers  can  compare 
different  age  groups  within  the  labour  force,  or  compare  recent  and  earlier  entrants  to  the 

' A number  of  other  regional  or  sectoral  studies  were  available  elsewhere  but  were  not  suitable  for  inclusion  in 
the  project  given  its  focus  on  national  systems. 


labour  market.  In  contrast,  national  school  leavers'  surveys  explicitly  take  account  of  the 
transition  process  by  examining  the  flow  out  of  education  into  the  labour  market  or  further 
education/training.  With  school  leavers’  survey  data,  one  has  the  advantage  of  being  capable 
of  observing  individual  trajectories  over  a certain  time-period  after  leaving  the  educational 
system.  That  is,  individual  labour  force  histories  are  directly  observed  as  a sequence  of  labour 
force  statuses  and  their  associated  features.  Obviously,  this  is  impossible  from  LFS  data 
sources,  as  the  necessary  information  is  simply  absent.  Still,  some  insights  into  transition 
processes  can  be  gained  from  generating  aggregate  career  paths  by  comparing  the 
distribution  of  labour  market  states  between  individuals  having  spent  different  amounts  of 
time  on  the  labour  market.  The  basic  difference  between  using  SLS  and  LFS  data  for 
transition  research  thus  lies  in  the  fact  that  SLS  allows  us  to  represent  individual  transition 
processes,  while  LFS  sources  are  restricted  to  analyses  of  aggregate  (average)  patterns  only.  It 
must  be  emphasised,  however,  that  the  two  data  sources  yield  complementary  insights  into  the 
transition  from  school  to  the  labour  market.  In  the  remainder  of  the  section,  we  discuss  these 
characteristics,  advantages  and  shortcomings  of  the  data  sources  in  greater  detail. 

3.4.1  Addressing  transition  processes  from  LFS  data 

The  EULFS  appears  as  an  attractive  database  to  comparative  research  for  a number  of 
reasons.  For  each  of  the  fifteen  current  member  states,  the  LFS  provides  information  based  on 
large  sample  sizes,  which  allow  for  differentiated  results  on  labour  force  activity  and  its 
determinants,  and  the  surveys  are  explicitly  administered  according  to  a design  which  is 
geared  towards  producing  comparable  information  across  countries.  In  addition,  the  LFS 
surveys  are  repeated  at  least  annually,  so  that  they  represent  one  of  the  few  databases  from 
which  current  processes  of  social  change  can  effectively  be  studied.  Last  but  not  least,  the 
LFS  in  general  contains  a wealth  of  information  on  current  labour  force  activities, 
employment  conditions  (occupation,  industry,  hours,  job  tenure  etc.)  and  socio-demographic 
characteristics  of  respondents.  These  advantages  of  the  EULFS  come  at  a price,  however,  at 
least  in  a study  on  school-to-work  transitions  (cf.  the  discussion  in  Cereq,  1997;  Couppie  and 
Mansuy,  1999). 

The  main  drawback  of  the  LFS  surveys,  at  least  as  currently  released  by  Eurostat,  is  their 
purely  cross-sectional  nature  at  the  level  of  individual  respondents.  That  is,  the  EULFS  by 
definition  does  not  allow  us  to  follow  a dynamic  research  approach  at  the  micro  level  which 


would  investigate  individual  trajectories  from  education  into  the  labour  market.  The  necessary 
information  is  simply  not  present  in  the  database  as  the  same  individuals  cannot  be  followed 
over  time.2  Still,  to  a limited  extent,  retrospective  information,  on,  for  example,  past 
employment  status,  is  available  from  LFS  sources.  In  fact,  this  information  has  been  used  in 
some  analyses  (Couppie  and  Mansuy,  2000a,  2000b),  and  some  results  from  these  will  be 
presented  below.  On  the  other  hand,  it  has  to  be  recognised  that  the  analytical  value  of  the 
retrospective  information  provided  is  in  itself  quite  limited:  as  retrospective  measures  on 
potential  explanatory  covariates  are  not  available,  causal  analysis  of  labour  market  flows  is 
greatly  inhibited  as  it  can  only  be  conducted  for  those  characteristics  which  are  (reasonably 
assumed  to  be)  stable  over  time.  In  addition,  measurement  concepts  for  labour  force  statuses 
are  not  identical  at  both  time  points,  so  that  definitional  problems  might  plague  any  such 
analysis.  Using  EULFS  data  for  transition  research  for  most  purposes  thus  has  to  represent  a 
conscious  decision  to  restrict  one’s  analytical  potential  to  what  is  available  from  cross- 
sectional  data  while  being  able  to  cover  all  EU  countries  in  the  research.  Consequently,  the 
LFS  analyses  performed  within  the  CATEWE  project  can  be  read  as  an  attempt  to  extract  as 
much  information  on  youth  labour  market  integration  in  Europe  as  possible  from  this  cross- 
sectional  database. 

The  key  to  our  analyses  is  to  realise  that,  although  it  is  impossible  to  observe  individual 
trajectories  between  education  and  work  in  the  LFS  data,  cross-national  similarities  and 
differences  in  macro-level  patterns  of  youth  labour  market  integration  can  readily  be  observed 
from  the  database.  In  fact,  as  information  is  collected  annually,  traditional  cohort  analyses  can 
be  performed  if  information  from  subsequent  surveys  is  linked  accordingly.  Muller  and 
Wolbers  (1999)  applied  this  technique  to  address  the  scope  of  educational  expansion  in 
Europe  over  the  past  decades:  by  combining  educational  distributions  for  the  same  birth 
cohorts  over  the  historical  observation  period  currently  available  from  the  EULFS,  they  have 
been  able  to  analytically  separate  life-cycle  patterns  of  educational  attainment  from  cohort 
effects  on  the  level  of  educational  attainment.  For  the  purpose  of  country  comparison,  this 
generates  a valid  picture  of  educational  attainment  processes  at  the  macro  level,  even  without 
the  availability  of  longitudinal  data  at  the  individual  level  and  without  the  imposition  of  any 
additional  assumptions  on  the  data. 

2 While  there  are  some  national  LFS  studies  which  actually  do,  and  others  which  at  least  in  principle  would 
permit  us  to  follow  this  approach,  the  harmonised  EULFS  does  not  so  far  allow  the  identification  of  members  of 
existing  rotating  samples  across  annual  survey  waves. 

A n 


O 

ERIC 


34 


This  approach  can,  in  principle,  also  be  applied  to  the  analysis  of  labour  market  outcomes  for 
those  entering  the  labour  force.  On  the  other  hand,  analyses  based  on  genuine  birth  cohorts 
are  unlikely  to  yield  adequate  (comparative)  representations  of  labour  market  entry  processes: 
within  any  single  country,  leavers  from  different  educational  backgrounds  exit  the  education 
and  training  system  at  different  ages.  That  is,  any  straightforward  differentiation  of  youth 
labour  market  outcomes  by  age  tends  to  misrepresent  the  situation  of  interest  as  young  people 
of  a given  age  might  actually  be  in  very  different  career  stages.  A university  graduate  might 
have  just  begun  working  in  her  first  job  at  age  27,  while  somebody  who  left  school  after 
attaining  his  compulsory  education  certificate  has  already  been  working  for  10  years.  The 
issue  becomes  even  more  problematic  in  comparative  research,  as  the  precise  biographical 
timing  of  these  transitions  varies  according  to  the  particular  institutional  structure  of  national 
education  and  training  systems.  Moreover,  differences  in  national  levels  of  educational 
attainment  can  generate  misleading  country  differences  in  macro-level  indicators  of  labour 
market  outcomes,  totally  unrelated  to  any  differences  in  actual  integration  processes.  For 
these  reasons,  most  of  the  project’s  analyses  of  labour  market  outcomes  are  based  on  labour 
force  entry  cohorts  rather  than  birth  cohorts.  That  is,  the  incidence  of  the  transition  period  has 
been  defined  relative  to  the  biographical  time  point  of  leaving  the  education  and  training 
system  rather  than  sheer  biological  age.  In  fact,  this  defines  our  perspective  on  the  transition 
period  as  one  on  labour  market  outcomes  in  the  early  career  stage,  after  having  completed 
initial  education  and  training.  In  most  analyses,  we  chose  typical  graduation  ages  for  different 
types  and  levels  of  education  as  published  in  the  OECD's  Education  at  a Glance  series  (see, 
for  example,  OECD,  1997)  as  an  approximation  to  the  biographical  time  point  of  completing 
a particular  type  of  education  in  the  various  European  countries,  and  calculate  a measure  of 
potential  labour  force  experience  on  that  basis.  For  the  purposes  of  the  project,  we  then  focus 
on  labour  market  outcomes  among  those  in  their  initial  years  on  the  labour  market,  that  is,  up 
to  five,  or  in  some  analyses  up  to  ten  or  even  fifteen  years  after  having  obtained  their  highest 
educational  qualification.  In  addition,  most  analyses  assume  relative  stability  in  the  structure 
of  transition  processes  in  the  short-run  so  as  to  enable  the  use  of  synthetic  cohort  approaches 
in  the  statistical  analyses.  By  doing  so,  we  are  able  to  extract  the  macro  level  properties  of 
transition  patterns  for  all  current  EU  member  states  from  LFS  databases.3 

3 Two  papers  actually  used  national  LFS  microdata  for  some  in-depth  analyses  of  particular  questions  in  a more 
restricted  sample  of  countries  (Brauns  et  al.,  1999;  lannelli  and  Soro  Bonmati,  2000),  but  the  methodological 
remarks  made  here  apply  to  these  analyses  as  well. 


Certainly,  this  still  represents  a serious  limitation  with  respect  to  an  adequate  description  of 
occurring  transition  processes.  It  is  most  discomforting  not  to  be  able  to  describe  individual 
level  transition  processes  because  the  full  extent  of  individual  heterogeneity  in  transition 
processes  between  education  and  work  cannot  possibly  be  uncovered  from  LFS  data.  On  the 
other  hand,  what  still  can  be  observed  from  this  database  under  the  chosen  set-up,  is  the 
average  labour  market  outcomes  of  young  people  in  their  early  career  stages  in  all  European 
countries  and  conditional  on  education,  time  since  leaving  education  and  training  systems, 
and  certain  socio-demographic  factors.  Exploring  the  incidence  of  cross-national  differences 
in  such  average  transition  outcomes  is  actually  a major  task  in  understanding  the  outcomes  of 
different  institutional  arrangements  regulating  school-to-work  transitions  in  Europe  which  has 
not  been  done  in  a similarly  encompassing  fashion  so  far. 

Of  course,  more  detailed  longitudinal  data  do  allow  for  more  detailed  studies  of  transition 
processes.  In  particular,  using  the  LFS  implies  the  restrictive  assumption  that  (a)  leaving  the 
education  and  training  system,  (b)  achieving  the  highest  qualification  and  (c)  entering  the 
labour  market  all  happen  at  the  same  time;  in  other  words,  that  only  the  most  recent  of 
potentially  multiple  entries  is  of  substantive  importance.  As  we  argue  in  3.7  below,  this  need 
not  be  the  case:  moreover,  the  relationship  between  these  three  events  is  variable  across 
countries.  Nevertheless  analyses  based  on  this  simplified  assumption  can  provide  useful 
insights  into  cross-national  differences,  which  are  consistent  with  project  research  based  on 
school  leaver  survey  data  as  well  as  previous  research,  as  the  work  summarised  in  section  3.5 
below  demonstrates. 

3.4.2  Analysing  transition  data  using  national  transition  surveys 

One  of  the  key  objectives  of  the  CATEWE  project  was  to  use  existing  national  transition 
surveys  to  explore  early  transition  processes  among  young  people  in  a range  of  countries.  In 
most  of  the  countries  concerned  (France,  Ireland,  the  Netherlands  and  Scotland)  these  surveys 
were  surveys  of  'school-leavers',  that  is,  those  exiting  secondary  education  at  a particular 
point  in  time,  although  the  precise  definition  of  a 'leaver'  can  vary  across  countries.4  However, 

4 In  the  Scottish  context,  for  example,  'leaver'  refers  to  those  leaving  the  general  secondary  school  system  with 
further  education  regarded  as  a destination.  Countries  also  vary  in  their  approach  to  apprentices  who  are 
differentially  treated  as  'leavers'  and  'labour  market  entrants'. 


in  the  Swedish  case,  data  related  to  a series  of  cohort  surveys  of  young  people  leaving 
education  at  various  stages.  National  transition  surveys  have  a number  of  advantages  in 
exploring  educational  outcomes  and  early  labour  market  experiences  among  young  people 
(see  Figure  3.3  and  Raffe,  2000).  Firstly,  they  tend  to  collect  detailed  information  on 
educational  background,  incorporating  dimensions  which  are  considered  important  in  the 
particular  institutional  context.  The  sensitivity  of  transition  surveys  to  the  national  context  is 
an  advantage  in  providing  a more  complete  view  of  (national  variations  in)  the  transition 
process.  Secondly,  they  allow  us  to  directly  relate  young  people's  educational  background  to 
their  experiences  in  labour  market  integration  at  an  individual  level.  Thirdly,  the  fact  that  they 
are  (for  the  most  part)  leavers'  surveys  means  that  we  are  looking  at  young  people,  most  of 
whom  entered  the  labour  market  at  the  same  time  and  therefore  searched  for  work  and  started 
their  careers  under  the  same  conditions.  Fourthly,  such  surveys  tend  to  provide  rich  data  on  a 
range  of  transition  'outcomes'  among  young  people,  covering  not  just  principal  activity  but 
different  dimensions  of  job  quality. 

National  transition  surveys  have,  however,  been  largely  under-exploited  in  cross-national 
analyses.  One  of  the  main  advances  made  by  the  CATEWE  project  has  been,  therefore,  to  use 
these  national  transition  surveys  to  construct  an  integrated  database  with  information  on  key 
aspects  of  young  people's  educational  and  labour  market  experiences.  In  fact,  a total  of  three 
databases  were  constructed  as  part  of  the  project: 

1.  A current  database.  This  was  based  on  the  most  recent  year  for  which  school  leavers' 

surveys  were  available.  This  database  covers: 

• France:  young  people  who  left  general  or  vocational  full-time  secondary  education 
(including  apprenticeships  undertaken  as  part  of  initial  education,  but  excluding 
General  Baccalaureat  and  agricultural  courses)  in  1993-4  and  who  did  not  continue  in 
full-time  education; 

• Ireland:  young  people  who  left  secondary  school  in  1995-6,  surveyed  in  autumn  1997. 
This  includes  those  who  left  Junior  or  Leaving  Certificate  and  Post-Leaving 
Certificate  courses.  Other  post-secondary  vocational  courses  count  as  destinations, 
along  with  apprenticeships,  training  schemes  and  third-level  education. 

• The  Netherlands:  young  people  who  left  secondary  education  (including  MBO)  in 
1995-6  and  did  not  enter  another  form  of  secondary  education.  They  were  surveyed  in 
autumn  1997.  In  this  survey,  apprenticeships  count  as  destinations. 

O 

ERIC 


37 


43 


• Scotland:  young  people  who  left  general  secondary  school  in  1993-4,  surveyed  in 
spring  1995.  Courses  in  further  education  colleges,  apprenticeships,  training  schemes 
and  higher  education  courses  count  as  destinations; 

• Sweden:  young  people  who  completed  lower-secondary  (compulsory)  education  in 
1993  and  were  surveyed  in  spring  1997. 

This  database  has  been  used  to  examine  cross-national  differences  at  a very  early  point  in 
the  transition  process,  one  to  one  and  a half  years  after  leaving  school. 

2.  A time-series  database.  Since  no  comparable  full  leavers'  survey  was  available  for  earlier 
years  for  France,  it  was  based  on  three  countries:  Ireland  (1980-1997;  5 time-points), 
Scotland  (1979-1995;  5 time-points),  Netherlands  (1989-1997;  3 time-points).  This 
database  has  been  used  in  order  to  explore  changes  in  institutional  and  labour  market 
contexts  over  time. 

3.  A longitudinal  database.  Since  the  current  and  time-series  databases  relate  to  a very  early 
point  in  young  people's  labour  market  career,  this  has  the  advantage  of  allowing  the 
analysis  of  longer  term  transition  patterns.  Unfortunately,  due  to  lack  of  data  availability, 
the  construction  of  a longitudinal  database  was  only  possible  for  Ireland  and  France,  and, 
for  a much  more  limited  set  of  variables,  for  France  and  Sweden. 

Constructing  the  integrated  databases  was  not  just  a means  of  conducting  substantive  analyses 
on  school  to  work  transitions  in  a number  of  countries  but  the  process  in  itself  yielded 
valuable  insights  into  our  understanding  of  institutional  variation  in  education,  training  and 
labour  market  systems.  The  procedure  involved  an  iterative  process,  involving  the 
identification  of  common  variables  within  the  national  datasets,  the  development  of  a 
'mapping'  from  the  original,  often  highly  diverse,  country-specific  variables  to  a common 
variable  specification  and  the  testing  of  these  new  comparative  variables  (for  further  details, 
see  CATEWE,  1999).  This  approach  yields  a number  of  advantages  over  the  analysis  of 
transition  surveys  at  a national  level  (see  Brannen,  Smyth,  2000;  CATEWE,  1999).  It  allows 
us  to  directly  test  cross-national  differences  in  transition  patterns,  controlling  for  a range  of 
other  factors.  Thus,  we  examine  whether,  for  example,  educational  relativities  in 
unemployment  differ  between  Ireland  and  the  Netherlands,  all  else  being  equal.  More 
importantly,  the  construction  of  comparable  variables  for  countries  with  very  different 
institutional  contexts  requires  a rigorous  clarification  of  the  different  dimensions  of  education 
and  transition  outcomes  explored  in  the  analyses.  Thus,  the  work  served  to  challenge  our  pre- 
O 

ERIC 


38 


A A 


existing  assumptions  about  the  nature  of  cross-national  variation  and  contributed  to  the 
development  of  new  classificatory  schema  for  analysing  different  dimensions  of  the  transition 
process.  In  particular,  the  construction  of  variables  specifically  for  the  purposes  of  the  project 
has  meant  that  we  can  directly  reflect  the  central  research  questions  we  seek  to  address  rather 
than  using  pre-existing  (and  often  inappropriate)  classification  systems.  It  has  also  helped  to 
develop  a set  of  multi-dimensional  indicators  which  better  reflect  the  specificities  of  the 
different  institutional  systems. 

However,  national  transition  surveys  and  any  integrated  database  drawing  on  these  surveys  do 
have  some  limitations.  They  cannot  allow  us  to  compare  the  experiences  of  young  people 
with  older  age-groups  or  with  those  who  entered  the  labour  market  at  a much  earlier  point  in 
time,  analyses  that  can  usefully  be  carried  out  using  the  Labour  Force  Surveys.  Thus,  it  can  be 
difficult  to  distinguish  whether  cross-national  differences  in  the  employment  experiences  of 
young  people  reflect  differences  in  the  labour  market  structures  as  a whole  or  in  the  relative 
position  of  young  workers  vis-a-vis  the  adult  population.  Furthermore,  differences  between 
the  national  surveys  in  design  and  content  result  in  difficulties  in  comparability  (see 
CATEWE,  1999).  These  difficulties  are  as  much  conceptual  as  technical  since,  in  the  context 
of  significant  institutional  differences  between  transition  systems,  it  is  impossible  to  identify  a 
single  transition  event  that  has  equivalent  significance  in  each  system  and  which  can  provide 
the  basis  for  comparison.  Finally,  such  surveys  are  available  only  for  a (limited)  number  of 
countries:  France,  Ireland,  the  Netherlands,  Scotland  and  Sweden.  Thus,  the  database  does  not 
include  any  'dual  system'  country  or  any  country  from  Southern  Europe,  groups  of  countries 
which  have  been  found  to  have  distinctive  profiles  in  terms  of  transition  processes  (see  Muller 
et  al.,  1999).  The  set  of  countries  studied  includes  one  country  often  grouped  with  the  dual 
system  countries  in  terms  of  the  predominance  of  occupational  labour  market  arrangements, 
the  Netherlands;  the  remainder  of  the  countries,  however,  come  from  the  group  of  North- 
Western  European  countries  usually  characterised  as  'ILM'  (internal  labour  market)  countries 
(see,  for  example,  Gangl,  1999).  This  distinction  has  also  been  characterised  as  the  difference 
between  systems  with  an  underlying  'employment  logic'  and  those  with  an  'education  logic' 
(Iannelli  and  Raffe,  2000).  However,  the  rich  data  from  the  national  surveys  allow  us  to 
explore  potential  heterogeneity  among  transition  systems  that  may  resemble  each  other  in 
other  respects. 


O 

ERLC 


39  45 


In  the  following  sections  of  the  chapter,  we  highlight  the  main  findings  of  the  analyses  using 
Labour  Force  Survey  and  national  school  leavers'  data. 

3.5  TRANSITIONS  FROM  EDUCATION  TO  WORK  IN  EUROPE  - LFS  RESULTS 

CATEWE  aimed  to  deliver  a genuine  European  perspective  on  transitions  from  education  to 
working  life.  This  promise  is  not  easily  fulfilled  as  adequate  longitudinal  data  from  which  to 
study  labour  market  integration  processes  at  the  individual  level  are  available  for  a limited 
subset  of  European  countries  only.  Therefore,  CATEWE  has  attempted  to  complement  its 
analyses  based  on  longitudinal  data  by  analyses  drawing  on  the  European  Union  Labour 
Force  Survey  (EULFS).  These  analyses  are  first  intended  to  provide  a broader  picture  on 
patterns  of  labour  market  entry  across  EU  countries,  including  those  where  longitudinal 
microdata  was  unavailable  to  the  project.  Analyses  of  EULFS  data  provide  a unique 
opportunity  to  situate  results  from  the  analysis  of  School  Leaver  Surveys  within  an  even 
broader  European  context.  But  apart  from  this  purpose,  the  project  also  attempted  to  make  use 
of  the  genuine  potential  of  the  EULFS  data  base  for  transition  research.  To  do  so  based  on 
cross-sectional  LFS  data  is  certainly  less  obvious  than  from  a truly  longitudinal  database,  but 
we  believe  that  the  EULFS  sources  have  some  inherent  qualities  of  their  own  in  that  respect, 
which  can  be  fruitfully  exploited  by  proper  statistical  analysis. 

Within  the  project,  ten  substantive  working  papers  suitable  for  later  scientific  publication 
have  been  produced  based  on  LFS  data  (see  Table  3.1  below).  Given  the  available 
information  in  LFS  sources,  most  of  the  papers  centre  around  the  education-employment 
linkage  in  European  countries.  Individual  papers  explore,  for  example,  cross-national 
similarities  and  differences  in  the  educational  background  of  young  people  entering  the  labour 
market,  as  well  as  the  nature  and  scope  of  educational  expansion  over  the  past  decades 
(Muller  and  Wolbers,  1999),  or  the  provision  of  dual  forms  of  vocational  training  in  European 
countries  and  their  evolution  over  the  last  decade  (Wolbers,  2000).  On  the  labour  market  side, 
there  are  papers  aimed  at  a broad  descriptive  overview  of  labour  market  outcomes  for  recent 
entrants  into  the  labour  force  (e.g.  Couppie  and  Mansuy,  2000a),  which  in  part  also  provide 
country  classifications  in  terms  of  relatively  similar  aggregate  transition  patterns  (Couppie 
and  Mansuy,  2000b;  Gangl,  2000a).  In  addition,  there  is  a set  of  more  analytical  papers  using 
advanced  multivariate  statistical  techniques  for  causal  analyses  of  unemployment  risks  and 

O 

ERIC 


40 


A 


employment  outcomes  among  labour  market  entrants  in  different  European  countries  (van  der 
Velden  and  Wolbers,  2000;  Gangl,  2000b,  2000c).  And  finally,  there  are  two  papers  which 
use  more  detailed  national  LFS  microdata  in  order  to  understand  the  peculiarities  of  transition 
outcomes  in  Southern  Europe  (Iannelli  and  Soro  Bonmati,  2000)  and  the  effects  of  education 
on  unemployment  processes  (Brauns  et  al.,  1999). 


Author(s) 

Hildegard  Brauns,  Markus 
Gangl,  and  Stefani  Scherer 
(1999) 


Thomas  Couppie  and 
Michele  Mansuy  (2000a) 

Thomas  Couppie  and 
Michele  Mansuy  (2000b) 


Markus  Gangl  (2000a) 


Markus  Gangl  (2000b) 


Markus  Gangl  (2000c) 


Cristina  Iannelli  and 
Asuncion  Soro  Bonmati 
(2000) 

Walter  Muller  and 
Maarten  Wolbers  (1999) 


Rolf  van  der  Velden  and 
Maarten  Wolbers  (2000) 

Maarten  Wolbers  (2000) 


Table  3.1 

Overview  of  LFS  Working  Papers 


Title 

Education  and  unemployment:  Patterns 
of  labour  market  entry  in  France,  the 
United  Kingdom,  and  West  Germany. 

The  Position  of  New  Entrants  on 
European  Labour  Markets. 

New  Entrants  and  experienced  workers 
on  European  Labour  Markets. 


European  Perspectives  on  Labour 
Market  Entry:  A Matter  of  Institutional 
Linkages  between  Training  Systems  and 
Labour  Markets? 

Education  and  Labour  Market  Entry 
across  Europe:  the  Impact  of 
Institutional  Arrangements  in  Training 
Systems  and  Labour  Markets. 

Changing  Labour  Markets  and  Early 
Career  Outcomes:  Labour  Market  Entry 
in  Europe  over  the  Past  Decade. 

The  Transition  from  School  to  Work  in 
Southern  Europe:  The  Cases  of  Italy 
and  Spain 

Educational  attainment  of  young  people 
in  the  European  Union:  cross-country 
variation  of  trends  over  time. 


The  integration  of  young  people  into  the 
labour  market  within  the  European 
Union:  the  role  of  institutional  settings. 

Learning  and  working:  Double  statuses 
in  youth  transitions  within  the  European 
Union. 


Main  Topics 

Role  of  education  for  avoiding  extensive 
periods  of  initial  job  search  and 
subsequent  job  instability  in  three 
European  countries 

Overview  of  labour  market  outcomes 
among  recent  entrants  to  European 
labour  markets 

Cross-national  similarities  and 
differences  in  various  aspects  of 
transition  patterns,  e.g.  dual  status 
situations,  unemployment  and 
employment  outcomes 

Cross-national  similarities  and 
differences  in  the  relations  between 
labour  force  experience,  qualifications 
and  unemployment  and  employment 
outcomes 

Cross-national  similarities  and 
differences  in  the  role  of  education  for 
unemployment  and  employment 
outcomes 

Effects  of  macroeconomic  and  macro- 
structural  trends  on  transition  outcomes 
in  Europe 

Comparison  of  the  patterns  of  transition 
from  education  to  the  labour  market  in 
Spain  and  Italy 

Cross-national  similarities  and 
differences  in  educational  backgrounds 
of  market  entrants  in  Europe;  similarities 
and  differences  in  the  nature  of 
educational  expansion 

Effects  of  institutional  context  factors  on 
unemployment  and  employment 
outcomes 

Cross-national  similarities  and 
differences  in  the  incidence  of  combined 
training  and  work  activities  in  Europe 


0 


The  following  represents  an  attempt  to  provide  a concise  summary  of  the  main  results  from 
our  analyses.  Rather  than  summarising  individual  papers,  this  review  will  be  organised  along 
the  substantive  themes  covered,  which  have  been  touched  upon  in  one  or  more  of  these 
analyses.  More  specifically,  project  results  are  reviewed  for  the  issues  of  (a)  educational 
achievement  and  the  nature  of  qualifications  among  school  leavers  in  Europe,  (b)  cross- 
national similarity  and  difference  among  European  countries  in  terms  of  transition  patterns 
from  education  to  work,  (c)  the  incidence  of  unemployment  among  market  entrants  and  the 
role  of  educational,  institutional,  and  other  contextual  determinants,  and  (d)  the  nature  of 
employment  outcomes  in  early  career  stage  and  their  determinants  across  European  countries. 

3.5.1  Educational  achievement  and  the  nature  of  qualifications  across  Europe 


In  a certain  sense,  education  is  the  key  resource  available  to  individuals  to  influence  their 
labour  market  fortunes.  Education  and  training  represents  an  individual  investment  in 
qualifications  which  are  afterwards  rewarded  on  the  labour  market.  This  general  relation 
already  bears  on  the  nature  of  education-to-work  transitions  as  different  types  of  training 
might  be  more  or  less  effective  in  generating  smooth  patterns  of  labour  market  entry.  To  the 
extent  that  training  systems  vary  between  countries,  the  different  resources  institutionally 
provided  to  young  people  can  be  expected  to  lead  to  substantially  different  transition  patterns 
in  different  European  societies.  And  indeed,  as  amply  shown  in  project  research  (notably 
Muller  and  Wolbers,  1999;  Wolbers,  2000),  the  qualificational  background  of  labour  market 
entrants  is  strikingly  different  across  European  Union  countries.  These  differences  relate  to 
both  the  level  of  educational  attainment  as  well  as  the  nature  of  qualifications  obtained.  More 
specifically,  three  broad  country  patterns  seem  to  emerge  from  our  data,  which  closely  reflect 
the  underlying  institutional  structures  in  education  and  training  systems  (Muller  and  Wolbers, 
1999). 


ERIC 


As  a first  type,  there  are  those  Continental  countries  operating  extensive  vocational  training 
systems  at  the  upper  secondary  level,  like  Austria,  Denmark,  Germany,  and  the  Netherlands, 
but  also  the  other  Nordic  countries  Sweden  and  Finland.  In  all  of  these  countries,  the 
proportion  of  young  people  not  progressing  beyond  compulsory  education  levels  is  very  low, 
typically  well  below  15  per  cent  of  a birth  cohort.  At  the  same  time,  a significant  proportion 
of  young  people,  typically  25  per  cent  of  a cohort  and  more  (with  the  exception  of  Austria), 
obtain  tertiary  level  qualifications.  But  the  most  distinctive  feature  is  the  fact  that  almost 


42 


A r\ 


everybody  who  left  the  educational  system  from  the  upper  secondary  level  will  have  obtained 
occupationally-specific  qualifications.  Of  course,  these  will  mainly  have  been  acquired  in  the 
context  of  dual  system  arrangements  in  Austria,  Germany,  and,  to  a lesser  extent,  in  Denmark, 
while  the  dominant  vocational  training  route  will  be  school-based  training  in  the  Netherlands, 
Sweden  and  Finland.  Compared  to  these  countries,  the  aggregate  pattern  of  educational 
attainment  is  somewhat  different  in  the  remaining  Northern  European  countries,  though. 
Broadly  speaking,  there  is  little  difference  between  Northern  European  countries  in  terms  of 
tertiary  level  graduation  rates.  The  UK,  and  more  so  Ireland,  France,  and  Belgium  differ  from 
the  former  set  of  countries  mainly  in  the  fact  that  fairly  large  proportions  of  upper  secondary 
level  leavers  enter  the  market  with  general  rather  than  vocational  qualifications.  In  addition, 
the  progression  beyond  compulsory  education  is  significantly  lower  in  these  countries  than  in 
the  other  Northern  European  countries.  In  fact,  Southern  Europe  constitutes  a third  empirical 
pattern,  mostly  distinguished  from  countries  like  the  UK  or  France  by  the  lower  level  of 
educational  attainment  (except  Spain),  rather  than  any  difference  in  the  vocational-general 
mix  at  the  upper  secondary  level.  If  anything,  then  Southern  European  education  and  training 
systems  provide  even  less  vocationally-specific  training  than  is  the  case  in  most  Northern 
European  countries.  As  an  illustration  to  these  distinctions,  Figure  3.4  below  depicts  the  life- 
cycle  pattern  of  educational  attainment  in  six  exemplary  European  countries  (Muller  and 
Wolbers,  1999). 


ERIC 


43 


4$ 


Figure  3.4 

Educational  Attainment  by  Age,  Selected  European  Countries 


Netherlands 


Age 


France 


United  Kingdom 


Source:  Muller  and  Wolbers,  1999. 

In  part,  these  country  differences  are  rapidly  changing.  As  the  analyses  of  Muller  and 
Wolbers  (1999)  show,  the  nature  and  pace  of  educational  expansion  has  varied  significantly 
between  European  countries  over  the  past  two  decades.  Those  countries,  for  example,  which 
previously  had  the  highest  proportions  of  individuals  with  only  compulsory  education  were 


also  the  most  successful  in  reducing  these  proportions  recently,  while  the  Nordic  countries  or 
Germany  and  Austria  have  been  much  less  able  to  reduce  these  figures  below  the  levels 
already  achieved  a generation  ago.  In  much  the  same  vein,  catching-up  processes  also 
occurred  at  the  higher  levels  of  education.  Recent  educational  expansion  of  tertiary  education 
has  been  occurring  fastest  in  Southern  European  countries,  notably  in  Spain  and  Portugal. 
Similarly  large  expansions  took  place  in  Ireland,  France  and  the  UK,  while  the  respective 
trends  have  been  much  weaker  in  Austria,  Germany  and  the  Nordic  countries.  In  sum,  there 
are  considerable  trends  under  way  towards  converging  educational  levels  among  young 
people  in  different  European  countries. 

While  the  levels  of  educational  achievement  may  actually  converge  somewhat  across 
European  countries,  it  is  much  less  likely  that  the  more  specific  nature  of  initial  qualifications 
provided  will  actually  converge  quickly.  As  has  been  indicated  above,  a crucial  distinction 
between  European  countries  is  the  extent  to  which  education  and  training  systems  already 
provide  occupationally-specific  training  (mostly  at  the  upper  secondary  level).  Most 
distinctive  to  systems  providing  occupationally-specific  training  are  large-scale  dual  system 
arrangements  as  operated  in  Austria,  Germany,  but  also  school-based  vocational  training  in 
the  Netherlands.  One  might  actually  argue  that  these  systems  are  merely  an  institutionally 
different  solution  for  providing  adequate  training  to  young  people.  In  cases  where  a dual 
system  exists,  training  provision  is  more  regulated  and  integrated  more  closely  into  the 
education  system,  while  in  countries  lacking  such  arrangements,  the  respective  training  is 
provided  by  companies  under  their  own  auspices. 


Results  from  the  project  cast  some  doubts  on  such  optimistic  perspectives,  however. 
Unsurprisingly,  an  analysis  by  Wolbers  (2000;  see  Table  3.2  below)  clearly  shows  that 
participation  in  dual  system  training  occurs  most  often  in  those  countries  operating  large-scale 
dual  systems.  To  a large  extent,  participants  come  from  compulsory  education  backgrounds, 
that  is,  they  participate  in  dual  system  training  as  a means  to  progress  beyond  the  lowest  level 
of  education.  Consistent  with  the  above  notion,  Wolbers  then  also  establishes  a slightly  higher 
tendency  for  Northern  Europeans  outside  the  core  occupationalised  systems  to  combine 
regular  employment  with  further  education.  But  apparently,  this  training  occurs  mostly  among 
tertiary  level  graduates  rather  than  the  lowest  qualified  — effectively,  it  is  thus  very  unlikely 
that  dual  system  training  foregone  is  made  up  by  company  training  for  the  least  qualified  later 
on.  That  is,  it  is  typically  not  those  leaving  from  compulsory  levels  of  education  who  receive 


ERIC 


K -f 


45 


subsequent  company  investments,  but  rather  those  individuals  who  already  bring  a high  level 
of  qualifications  to  the  work  place.  To  do  full  justice  to  Wolbers'  results,  one  should  also  note 
that  precisely  those  Northern  European  countries  lacking  large  scale  vocational  training 
arrangements  have  been  those  which  significantly  extended  the  provision  of  training  which 
combines  learning  and  working.  Neither  in  the  traditional  occupationalised  systems  nor  in 
Southern  Europe  did  the  proportion  of  young  people  receiving  such  training  change 
substantially  over  the  last  decade.  In  Southern  Europe,  in  particular,  the  likelihood  of 
receiving  occupationally  relevant  training  after  leaving  the  education  and  training  system  is 
very  low. 


Table  3.2: 

The  Structure  of  Combined  Work-Training  Activities,  by  Institutional  Contexts 


Study 

Macro-institutional  Context 

Effects 

Wolbers,  2000 

Dual  System  Countries  (incl.  NL) 

highest  probability  of  dual  system  training,  stable  over 
time;  less  strong  gender-typing  of  dual  system  training 

Southern  Europe 

lowest  probability  of  dual  system  training,  working 
students  and  further  education  among  employed 

other  European  countries 

relatively  low  probability  of  dual  system  training,  but 
increasing  over  time;  highest  probability  of  further 
education  among  employed 

3.5.2  Cross-national  similarity  and  difference  in  transition  outcomes:  a broad 
perspective  on  labour  market  entry  patterns  in  European  countries 


Conditional  on  leaving  the  education  and  training  system,  what  are  the  similarities  and 
differences  between  European  countries  in  terms  of  labour  market  outcomes  which  occur  on 
entering  the  labour  force?  And  if  differences  occur,  are  there  some  countries  which  are 
relatively  similar  to  each  other  in  terms  of  the  observed  outcome  patterns  while  others  differ? 
Which  contours  does  a European  map  of  transition  experiences  show?  In  fact,  these  questions 
can  be  answered  along  numerous  dimensions,  each  emphasising  a particular  aspect  of  labour 
market  and  employment  outcomes.  Within  the  project,  most  research  has  focused  on 
unemployment  on  the  one  hand  and  job  features  like  occupation,  industry,  and  type  of 
contract  on  the  other,  although  alternative  measures  have  also  been  considered  to  some  extent 
(Couppie  and  Mansuy,  2000a,  2000b;  Gangl,  2000a).  In  addition,  these  analyses  attempted  to 
arrive  at  a descriptive  account  of  major  similarities  and  differences  in  transition  patterns 

ERJC 


46 


52 


between  European  countries,  often  yielding  empirical  classifications  of  countries  according  to 
observed  similar  transition  patterns.  Within  the  project,  several  approaches  have  been 
followed,  focusing  either  more  on  cross-national  variation  in  outcome  distributions  like 
proportions  of  dual  system  training,  youth  unemployment  rates  or  average  occupational  status 
outcomes  (Couppie  and  Mansuy,  2000b),  or  more  on  cross-national  variation  in  the  relations 
between  qualifications,  labour  force  experience  and  employment  outcomes  (Couppie  and 
Mansuy,  2000a;  Gangl,  2000a).  In  doing  so,  attention  was  given  to  both  comparative 
perspectives  on  the  features  of  national  youth  labour  markets  and  the  relationships  between 
youth  and  adult  labour  markets. 

In  fact,  there  are  some  broad  cross-national  similarities  between  European  countries  in  terms 
of  labour  market  experiences  among  recent  entrants  to  the  market,  notably  as  compared  to 
those  of  more  experienced  workers  (Couppie  and  Mansuy,  2000a,  2000b;  Gangl,  2000a). 
Typically,  unemployment  rates  are  higher  at  the  early  career  stages  as  people  have  to  look  for 
a first  job  or  have  been  able  to  secure  only  fairly  uncertain  jobs  in  the  beginning  (Couppie  and 
Mansuy,  2000a,  2000b;  Gangl,  2000a).  Similarly,  those  entering  the  market  are 
disproportionately  allocated  to  low-skilled  service-sector  jobs  (Couppie  and  Mansuy,  2000a, 
2000b;  Gangl,  2000a),  often  under  fairly  precarious  contract  conditions,  as  signified,  for 
example,  by  the  higher  incidence  of  fixed-term  contracts  among  market  entrants  (Couppie  and 
Mansuy,  2000a).  Moreover,  transitions  between  labour  market  statuses  of  employment, 
unemployment,  and  inactivity  occur  much  more  often  among  market  entrants  compared  to 
more  experienced  workers  (Couppie  and  Mansuy,  2000a). 


O 

ERIC 


47 


53 


Figure  3.5 

Unemployment  Rates  and  Labour  Force  Experience 


'Austria  *0—  Denmark 

■Germany  -^r-1  Netherlands 


—O—  Belgium 

France 

•■^“Ireland 

—^—United  Kingdom 

-O-Greece  ""O— Italy  | 
^^"Portugal  Spain  | 


Notes:  Leavers  from  ISCED  level  3;  lines  represent  results  from  logarithmic  smoothing. 
Source:  Gangl,  2000a. 


On  the  other  hand,  it  is  important  to  recognise  that  national  transition  patterns  are  far  from 
identical,  even  if  some  aspects  are  common  to  most,  if  not  all,  of  them.  Indeed,  countries 
differ  markedly  with  respect  to  some  core  aspects  of  youth  transition  experiences.  There  are 
some  countries  where  unemployment  risks  among  market  entrants  are  markedly  more 
pronounced  than  those  for  more  experienced  workers.  The  Southern  European  countries,  but 
also  France,  are  examples  for  these  (cf.  Figure  3.5  above  as  an  illustration  for  the  group  of 
leavers  from  upper  secondary  (ISCED  level  3)  education).  But  there  are  also  other  countries 
where  this  relationship  is  extremely  weak,  so  that  unemployment  rates  among  market  entrants 
closely  parallel  those  among  more  experienced  workers.  Austria,  Denmark,  Germany,  and  the 
Netherlands  would  be  examples  of  the  latter  group  of  countries  (Couppie  and  Mansuy,  2000a, 
2000b;  Gangl,  2000a).  In  much  the  same  way,  the  degree  of  disproportionate  allocation  to 
lower-level  employment  or  to  the  service  sector  varies  between  European  countries:  while 
many  young  people  enter  the  labour  market  at  particularly  low  job  levels  and  then  progress 
over  their  initial  years  in  the  labour  market  in  terms  of  occupational  status  or  similar  measures 
of  job  characteristics,  this  tendency  is  significantly  weaker  in  Germany  or  Austria,  for 
example  (Couppie  and  Mansuy,  2000a;  Gangl,  2000a).  Also,  it  seems  relatively  common 

ERIC 


48 


K.  A 


among  European  countries,  that  a considerable  proportion  of  new  entrants  to  the  market  enter 
non-standard  forms  of  employment,  which  are  then  increasingly  left  over  the  initial  years  in 
the  labour  force.  Typically,  some  20  per  cent  of  an  entry  cohort  held  temporary  contracts  in 
their  first  year  on  the  market,  although  the  Nordic  countries,  and  even  more  so  Spain  exhibit 
markedly  higher  figures,  with  estimates  ranging  even  up  to  80  per  cent  in  the  Spanish  case 
(Couppie  and  Mansuy,  2000a).  In  many  respects,  similar  patterns  are  also  evident  in  the  case 
of  part-time  employment.  While  part-time  employment  in  general  is  much  less  specific  to  the 
early  career  stage,  Couppie  and  Mansuy  (2000a)  show  that  the  incidence  of  involuntary  part- 
time  contracts  clearly  declines  with  increasing  labour  force  experience  in  almost  all  European 
countries.  But  again,  countries  differ  remarkably  in  the  extent  to  which  young  people  have  to 
accept  involuntary  part-time  employment:  notably  in  Belgium,  France,  Sweden  and  Finland 
the  respective  proportions  amounted  to  well  above  10  per  cent  among  new  entrants  to  the 
labour  market. 

These  findings  of  important  heterogeneity  among  the  countries  also  extend  to  particular  types 
of  labour  market  mobility.  While  it  is  true  for  many  countries  that  young  people  are  faster  to 
leave  unemployment,  there  are  important  exceptions  to  this  rule.  In  neither  Italy  nor  Greece 
does  the  likelihood  of  leaving  unemployment  vary  by  experience,  and  the  same  holds  for  the 
UK;  at  the  same  time,  the  transition  rates  from  unemployment  to  employment  in  the  UK  are 
about  twice  those  for  Greece  and  Italy  (Couppie  and  Mansuy,  2000a;  cf.  Figures  3.6  and  3.7 
below).  Similar  observations  can  be  made  with  respect  to  the  probability  of  losing 
employment  and  subsequently  entering  unemployment,  where  most  countries  exhibit  a 
modestly  negative  relationship  with  increasing  work  experience.  Spain  and  France,  in 
particular,  experience  excessively  high  inflow  rates  among  market  entrants,  however 
(Couppie  and  Mansuy,  2000a;  cf.  Figure  3.6  below).  Behind  all  these  descriptive  findings,  the 
main  substantive  result  is  to  realise  that  European  countries  differ  much  less  in  terms  of 
labour  market  outcomes  among  experienced  workers  than  they  do  in  terms  of  outcomes 
among  market  entrants.  To  the  extent  labour  market  entrants'  fortunes  differ  across  European 
countries,  this  reflects  cross-national  variation  in  the  relative  competitiveness  of  those  leaving 
the  education  and  training  systems,  that  is,  the  extent  to  which  market  entrants  achieve  similar 
outcomes  as  experienced  workers  along  a number  of  dimensions.  Variation  in  this 
relationship  is  at  the  core  of  empirically  distinguishable  'transition  systems'  among  European 
countries. 


Figure  3.6:  Job  exit  rates  among  individuals  employed  in  the  previous  year,  by  labour  force  experience 


r- 

m 


\n 


© 

IT) 


Sources:  Couppie  and  Mansuy,  2000a 


The  case  for  institutional  explanations  of  market  entrants'  competitiveness  in  different 
countries  could  be  strengthened  if  one  were  able  to  show  that  countries  exhibiting  similar 
institutional  frameworks  in  training  systems  and  labour  markets  are  actually  relatively  similar 
in  such  overall  transition  patterns.  In  our  analyses,  we  have  attempted  to  demonstrate  this  by 
exploring  the  nature  of  cross-national  similarities  and  differences  on  a number  of  labour 
market  dimensions  from  cluster  analyses  (Couppie  and  Mansuy,  2000b;  Gangl,  2000a).  And 
although  the  technique  is  exploratory  in  nature,  it  does  show  some  intriguing  profiles  of 
European  differences  in  transition  patterns.  Our  different  analyses  reliably  singled  out  two 
polar  transition  patterns  deviating  clearly  from  the  rest  of  Europe:  the  occupationalised 
systems  of  Austria,  Denmark,  the  Netherlands,  and  Germany,  on  the  one  hand,  and  the 
Southern  European  countries,  including  Italy,  Greece,  and  Portugal  at  least,  on  the  other. 

The  main  features  of  the  first  ideal-typical  pattern  was  that,  by  and  large,  market  entrants 
achieve  only  slightly  less  favourable  market  outcomes  than  more  experienced  workers,  in 
terms  of  both  unemployment  and  job  characteristics.  In  contrast,  in  countries  like  the  UK, 
France  and  Ireland,  market  entrants  are  significantly  disadvantaged  compared  to  more 
experienced  workers  on  both  dimensions.  The  Southern  countries,  in  turn,  deviate  from  this 
pattern  by  even  more  marked  disadvantages  to  market  entrants  in  terms  of  unemployment 
risks,  but  even  more  so  in  the  low  level  of  mobility  between  labour  market  statuses,  once 
initial  employment  has  been  secured.  Labour  markets  in  Northern  European  countries  exhibit 
much  larger  mobility  rates  between  employers  and  between  employment  and  unemployment 
than  is  the  case  in  the  typical  Southern  European  experience.  In  fact,  this  criterion  yields  a 
major  reason  for  considering  youth  experiences  in,  for  example,  Spain  as  relatively  similar  to 
France  rather  than  to  Portugal. 

Compared  to  the  polar  cases,  further  divisions  among  the  remaining  European  countries 
emerged  less  clearly.  It  is  clear  that  the  remaining  countries  comprising  France,  the  UK, 
Belgium,  Ireland,  Spain,  but  possibly  also  Sweden  and  Finland,  form  a much  less 
homogeneous  set  of  transition  profiles  than  those  described  earlier.  It  is  certainly  possible  to 
draw  finer  distinctions  among  transition  patterns  in  these  countries,  but  pur  research  based  on 
LFS  data  has  so  far  not  given  definite  results.  Depending  very  much  on  the  particular 
indicators  considered,  sometimes  Britain  and  Ireland  could  be  distinguished  from  a particular 
French  pattern,  but  in  other  analyses  Ireland  became  included  among  the  Southern  European 
O 

ERIC 


51  58 


countries.  Clearly,  this  uncertainty  in  the  results  reflects,  to  a large  part,  the  exploratory 
nature  of  the  particular  methodology  applied  and  differences  between  the  analyses  in  terms  of 
the  precise  indicators  utilised.  But  in  fact,  as  some  fairly  clear-cut  broad  types  of  different 
national  transition  patterns  seem  to  characterise  the  overall  European  experience,  there 
appears  some  scope  for  institutionally-based  explanations  of  these  patterns.  It  is  to  these  to 
which  we  now  turn. 

3.5.3  Unemployment  risks:  education,  institutions,  and  socio-economic  context 
conditions 


Among  the  various  aspects  of  school-to-work  transitions  of  potential  interest,  the  project’s 
more  specific  and  most  sophisticated  analyses  have  focused  on  two  core  labour  market 
outcomes,  namely  unemployment  risks  and  employment  outcomes  among  labour  market 
entrants.  In  the  respective  analyses,  we  attempted  to  explain  these  transition  outcomes  by 
adequately  accounting  for  both  the  role  of  individual  resources  and  characteristics  and  the 
impact  of  particular  institutional  contexts  and  other  macrostructural  and  socio-economic 
context  conditions  (van  der  Velden  and  Wolbers,  2000;  Gangl,  2000b,  2000c).  In  order  to 
properly  accomplish  this  task,  the  three  relevant  papers  applied  several  variants  of  multilevel 
analysis  as  a methodological  innovation  in  comparative  empirical  research.  While  this  section 
will  summarise  our  results  with  respect  to  determinants  of  unemployment  risks,  the 
subsequent  section  will  discuss  the  determinants  of  specific  employment  outcomes  in  greater 
detail. 


ERIC 


Among  the  many  factors  which  could  potentially  be  linked  to  the  incidence  of  unemployment, 
our  results  mainly  concern  three  types  of  determinants:  individual  education  and  training 
(Gangl,  2000b;  Brauns  et  al.,  1999),  institutional  features  of  both  education  and  training 
systems  and  labour  markets  (van  der  Velden  and  Wolbers,  2000;  Gangl,  2000b),  and  socio- 
economic context  conditions  (Gangl,  2000c).  Unsurprisingly,  each  factor  turns  out  to  have 
important  consequences  for  the  extent  of  labour  market  integration  problems  among  recent 
entrants.  Education  and  training,  for  example,  is  the  primary  individual  resource  for  avoiding 
unemployment  at  entering  the  labour  market.  In  general  and  controlling  for  other  factors,  the 
higher  the  individual  level  of  education  attained,  the  lower  the  risk  of  unemployment 
incidence  in  the  early  career  stage.  At  the  same  time,  vocational  training,  notably  if  obtained 


52 


59 


in  the  context  of  dual  training  arrangements,  also  contributes  to  lower  unemployment  risks 
(van  der  Velden  and  Wolbers,  2000;  Gangl,  2000b).  Figure  3.8  below  represents  these 
relationships  graphically,  based  on  a multilevel  model  which  controls  for  other  individual 
factors,  as  well  as  institutional  and  economic  context  factors. 

However,  the  interesting  result  is  that  this  relationship  does  not  hold  in  all  European 
countries;  rather,  there  is  systematic  institutional  variation  as  to  whether  educational 
credentials  serve  to  lower  the  individual  risk  of  unemployment.  In  fact,  there  is  little  variation 
among,  broadly  speaking,  Northern  European  countries,  except  for  the  better  performance  of 
those  who  achieve  school-based  vocational  training  in  occupationalised  labour  market 
contexts  like  Germany  and  the  Netherlands.  That  is,  apprenticeships  and  similar  types  of  dual 
system  training  lead  to  lower  unemployment  rates  than  those  of  upper  secondary  general 
tracks  in  both  occupationalised  and  less  occupationally  structured  systems.  The  difference 
between  these  two  types  of  transition  systems  lies  in  the  fact  that  leavers  from  school-based 
vocational  training  face  lower  unemployment  rates  in  more  occupationalised  contexts.  This 
finding  seems  to  support  the  reasoning  that  appropriate  vocational  specialisation  is  important 
to  integrate  young  people  into  the  labour  force  in  markets  exhibiting  strong  occupational 
boundaries  (irrespective  of  whether  the  qualification  is  obtained  from  school-based  or  dual 
forms  of  training),  while  in  less  tightly  structured  systems  it  is  more  the  actual  training 
contract  with  a particular  employer  (as  an  apprentice  or  otherwise)  which  reduces  subsequent 
unemployment  risks. 

But  the  main  institutional  divergence  occurs  in  Southern  European  countries,  where  the  level 
and  type  of  education  hardly  affects  unemployment  risks  at  all.  That  is,  while  low  qualified 
school  leavers  do  not  face  particularly  different  unemployment  risks  compared  to  their 
Northern  European  counterparts,  unemployment  rates  among  leavers  from  upper  secondary 
education  and  even  among  university  graduates  are  at  similar  levels  to,  rather  than 
substantially  lower  than,  those  among  the  least  qualified.  In  contrast  to  Northern  Europe, 
unemployment  in  the  early  career  stages  is  a particular  problem  of  the  highly  qualified  in  the 
South  rather  than  among  the.  least  qualified.  In  fact,  the  nature  of  unemployment  itself  is  thus 
likely  to  be  very  different  between  Northern  and  Southern  Europe. 


O 

ERIC 


53 


60 


Figure  3.8 

Unemployment  among  Market  Entrants:  Effects  of  Education  and  Institutional  Context 


Lower  secondary  Apprenticeships  Upper  secondary  Upper  secondary  Lower  tertiary  University  degree 

vocational  general 


Notes:  Predicted  probabilities  at  mean  individual  covariates  and  macrolevel  context  conditions,  based  on 

multilevel  regression  estimation. 

Source:  Gangl,  2000b 

In  order  to  deepen  our  understanding  about  the  role  of  education  in  actual  unemployment 
processes  underlying  the  above  results,  we  have  conducted  a more  sophisticated  analysis  of 
educational  effects  on  unemployment  processes  based  on  LFS  microdata  for  France,  the 
United  Kingdom,  and  Germany  (Brauns  et  al.,  1999).  In  that  analysis,  we  estimated  a two- 
stage  model  of  the  labour  market  entry  process,  distinguishing  between  unemployment  risks 
due  to  prolonged  initial  job  search  and  unemployment  risks  related  to  the  instability  of  initial 
employment  found.  From  this  analysis,  it  appeared  that  educational  resources  have 
reinforcing  effects  on  both  stages,  that  is,  those  qualifications  which  provide  relatively 
smooth  access  to  first  jobs  also  typically  provide  access  to  more  stable  employment.  This 
applies  in  particular  to  apprenticeships,  which  are  found  to  provide  not  only  almost  immediate 
access  to  employment  (e.g.  by  continued  employment  in  the  training  firm),  but  also  relatively 
secure  first  jobs. 


Apart  from  this  different  role  of  education  in  labour  market  allocation  processes,  the 
institutional  structure  of  education  and  training  systems  actually  exerts  a crucial  influence  in 


0 

ERIC 


54 


61 


itself  as  it  determines  the  qualificational  resources  available  to  labour  market  entrants  to  a 
large  extent.  Those  countries  operating  large-scale  dual  systems  of  training  provision 
experience  significantly  lower  unemployment  rates  in  the  transition  period  because  a large 
proportion  of  those  leaving  the  education  and  training  system  have  acquired  a qualification 
which  implies  direct  access  to  subsequent  employment  (van  der  Velden  and  Wolbers,  2000; 
Gangl,  2000b).  According  to  estimates  from  Gangl  (2000b),  this  effect  alone  amounts  to 
lower  aggregate  unemployment  rates  among  labour  market  entrants  in  dual  system  countries 
(including  the  Netherlands)  by  5 percentage  points  as  compared  to  the  other  Northern 
European  countries.  In  addition,  there  is  also  a composition  effect  of  educational  levels  on 
unemployment  rates,  which  further  disadvantages  the  Southern  European  countries  as 
compared  to  Northern  Europe:  the  higher  the  level  of  education  among  market  entrants,  the 
lower  a country's  unemployment  rate  among  this  group.  Beyond  these  institutional  effects  of 
educational  systems  and  broad  labour  market  contexts,  there  is  little  evidence  for  other 
relevant  institutional  factors.  Van  der  Velden  and  Wolbers  (2000)  tested  for  effects  of  a 
number  of  institutional  features  of  labour  markets,  including  wage  bargaining  structures, 
union  density,  the  extent  of  youth  activation  and  training  measures  among  others,  but  none  of 
these  receives  clear  empirical  support.  Only  in  the  case  of  the  strictness  of  employment 
protection  legislation,  they  find  evidence  for  a small  positive  effect  on  unemployment.  That 
is,  the  better  protected  the  core  work  force,  the  more  difficult  it  is  for  youth  to  successfully 
compete  in  securing  employment.  On  the  other  hand,  this  effect  did  not  receive  clear  support 
as  soon  as  the  structure  of  training  systems  was  simultaneously  controlled  for. 

Table  3.3 


Institutional  Effects  On  Unemployment  Rates  among  Market_  Entrants^ 


Study 

Institutional  Variable 

Effects 

van  der  Velden  / 
Wolbers,  2000 

Centralised  Wage  Bargaining 
Union  Density 

negative,  not  significant 
no  effect 

Employment  Protection 

positive,  not  significant 

Vocational  Specificity  / Educ. 

no  effect 

Dual  System 

negative 

Tracking  / Second.  Educ.  System 

no  effect 

Gangl,  2000b 

Apprenticeship  Systems 

negative 

Occupationalized  Markets 

negative  (for  vocational  qualifications) 

Southern  Europe 

positive  (for  better  qualified) 

Gangl,  2000c 

Interaction  of  macroeconomic  trends  and 
three  macroinstitutional  contexts 

cyclical  effects  less  pronounced  in  Southern 
Europe 

O 

ERIC 


55 


62 


At  the  same  time,  young  people's  labour  market  fortunes  are  not  isolated  from  the  evolution 
of  the  labour  market  in  general.  It  is  not  only  individual  qualifications  and  national 
institutional  contexts  which  affect  unemployment  risks  in  early  career  stages,  but  obviously 
also  the  broader  structural  context.  The  role  of  aggregate  macroeconomic  conditions, 
measured  by  either  aggregate  unemployment  rates  or  employment  growth  rates,  is  a key 
determinant  of  unemployment  risks  among  recent  entrants  to  the  labour  market  (van  der 
Velden  and  Wolbers,  2000;  Gangl,  2000b,  2000c).  Those  in  their  early  career  stages  are 
particularly  affected  by  cyclical  market  swings  as  they  are  typically  among  the  less 
competitive  individuals  on  the  market  and  have  not  yet  entered  stable  permanent  job 
positions.  In  each  recessionary  period,  unemployment  rates  among  market  entrants  increase 
relatively  stronger  than  aggregate  rates,  but  they  also  decline  more  strongly  in  more  buoyant 
times.  As  more  detailed  analyses  show,  it  is  the  lowest  qualified  school  leavers  whose  labour 
market  chances  are  particularly  vulnerable  to  cyclical  macroeconomic  developments  (Gangl, 
2000c).  But  there  is  yet  another  important  reason  why  those  entering  the  labour  market  with 
low  qualifications  form  a particular  problem  group.  According  to  the  results  in  Gangl  (2000c), 
ongoing  professionalisation  of  the  labour  force  and  related  increases  in  skill  requirements 
increasingly  work  against  the  lowest  qualified  school  leavers.  In  addition,  there  is  no  evidence 
that  any  European  country  is  exempted  from  this  tendency. 


3.5.4  Types  of  jobs  and  the  nature  of  employment  contracts:  some  determinants 


Understanding  unemployment  risks  is  one  important  element  in  understanding  transition 
processes,  yet  the  flip  side  of  the  coin  is  to  understand  young  people’s  employment  outcomes 
in  their  early  career  stages.  Several  of  the  project  analyses  have  touched  on  these  matters  by 
addressing  the  nature  of  occupational  allocation  of  labour  market  entrants  (Gangl,  2000b, 
2000c)  and  the  types  of  contracts  obtained  initially  (van  der  Velden  and  Wolbers,  2000).  We 
have  not  conducted  any  analysis  on  wage  or  earnings  outcomes  as  no  measures  on  them  are 
provided  in  the  EULFS,  at  least  up  to  the  1997  wave  which  has  been  the  most  recent  one 
considered  in  our  work.  And  as  in  the  case  of  unemployment,  our  main  analytical  interests 
focused  on  the  role  of  education  and  training,  institutions  and  socio-economic  context  factors 
in  generating  job  outcomes  at  the  start  of  individual  careers. 


O 

ERIC 


56 


In  fact,  the  role  of  education  and  training  as  a major  individual  resource  in  job  competition 
emerges  very  clearly  from  our  analyses.  The  higher  the  level  of  education  attained,  the  higher 
the  occupational  status  of  job  positions  (Gangl,  2000b),  the  lower  the  likelihood  of  entering 
into  low-skilled  jobs  (Gangl,  2000b),  the  higher  the  probability  of  accessing  professional  job 
positions  already  in  the  early  career  stage  (Gangl,  2000b),  the  lower  the  probability  of 
obtaining  fixed-term  or  otherwise  temporary  job  contracts  (van  der  Velden  and  Wolbers, 
2000),  and  the  higher  the  likelihood  of  having  a full-time  contract  (van  der  Velden  and 
Wolbers,  2000).  And  more  specifically,  it  turns  out  that  some  of  the  particular  advantages  of 
apprenticeship  contracts  in  terms  of  unemployment  risks  come  at  the  expense  of  allocation  to 
lower  level  jobs:  compared  to  leavers  from  general  or  school-based  vocational  tracks  at  the 
upper  secondary  level,  apprentices  attain  employment  in  lower  status  occupations  and  run  a 
higher  risk  of  entering  low-skilled  jobs.  In  fact,  there  are  few  indications  that  these 
relationships  vary  dramatically  between  the  various  European  countries:  in  general,  macro- 
institutional  differences  play  a much  more  limited  role  with  respect  to  employment  outcomes 
among  labour  market  entrants  than  is  the  case  with  respect  to  unemployment  risks  in  the  early 
career  stages. 

Again,  there  is  a certain  role  to  play  on  the  part  of  the  institutional  structure  of  education  and 
training  systems.  In  an  almost  trivial  sense,  the  higher  the  level  of  educational  attainment  in  a 
cohort  entering  the  labour  market,  the  higher  will  be  the  level  of  jobs  for  which  they  compete. 
The  still  lower  educational  levels  in  Southern  Europe  explain  the,  on  average,  lower 
occupational  attainment  levels  in  the  early  career  stages  there  to  a good  deal  already  (Gangl, 
2000b).  But  potentially  more  interesting  are  the  favourable  effects  of  large  scale  dual  systems 
or  similar  forms  of  vocationally  specific  training  provision.  According  to  our  results,  the 
presence  of  such  systems  lowers  the  incidence  of  low-skilled  employment  (Gangl,  2000b)  and 
temporary  contracts  (van  der  Velden  and  Wolbers,  2000)  among  young  people  entering  the 
labour  market.  To  which  extent  this  effect  is  due  to  the  occupational  specificity  of  the  training 
provided  itself  or  to  the  fact  that  the  offered  training  tracks  represent  a low-threshold  option 
for  attaining  education  and  training  beyond  compulsory  levels  is  an  open  question  to  future 
research  - but  the  effect  itself  is  undeniably  there. 


ERIC 


57  6 4 


Table  3.4 


Institutional  Effects  On  Employment  Outcomes  among  Market  Entrants 


Study  / 

Dependent  Variable 

Institutional  Variable 

Effects 

van  der  Velden  / 

Centralised  Wage  Bargaining 

negative,  not  significant 

Wolbers,  2000: 

Union  Density 

no  effect 

Temporary  Contract 

positive 

Employment  Protection 

Vocational  Specificity  / Educ. 

no  effect 

Dual  System 

negative 

Tracking  / Second.  Educ.  System 

no  effect 

van  der  Velden  / 

Centralised  Wage  Bargaining 

positive,  not  significant 

Wolbers,  2000: 
Part-Time 

Union  Density 

no  effect 

Employment 

Employment  Protection 

negative,  not  significant 

Vocational  Specificity  / Educ. 

positive 

Dual  System 

negative,  not  significant 

Tracking  / Second.  Educ.  System 

negative,  not  significant 

Gangl,  2000b: 

Apprenticeship  Systems 

small  negative  effect 

Occupational  Status 

Occupationalised  Markets 

no  effect 

Southern  Europe 

no  effect 

Gangl,  2000b: 

Apprenticeship  Systems 

negative  for  secondary  sector  employment 

Occupational 

Segment 

Occupationalized  Markets 

positive  for  professional  employment 

Southern  Europe 

no  effect 

Gangl,  2000c: 

Interaction  of  macroeconomic  trends  and 

positive  effects  of  professionalisation 

Occupational  Status 

three  macro-institutional  contexts 

strongest  in  OLM  countries,  negative  effects 
of  educational  expansion  strongest  in 
Southern  Europe 

Gangl,  2000c: 

Interaction  of  macroeconomic  trends  and 

positive  effects  of  professionalisation 

Occupational 

Segment 

three  macro-institutional  contexts 

strongest  in  OLM  countries 

Apart  from  this,  there  is  also  some  evidence  for  slightly  different  allocation  mechanisms 
operating  in  the  occupationalised  markets  of  Austria,  Denmark,  Germany,  and  the 
Netherlands.  In  these  systems,  occupational  allocation  tends  to  be  more  strongly  skill-based: 
as  job  competition  relies  more  heavily  on  (formally  certified)  skills  rather  than  experience, 
those  entering  the  labour  market  in  such  contexts  are  relatively  more  competitive  to  adult 
workers  than  is  the  case  in  systems  less  reliant  on  certified  skills.  Hence,  occupational  and 
employment  outcomes  reflect  more  adequate  matches  at  earlier  career  stages  than  elsewhere. 
In  support  of  this  reasoning,  Gangl  (2000b)  provides  evidence  that  higher  levels  of  education 
provide  more  protection  from  entering  low-skilled  jobs  in  occupationalised  systems,  and  that 
leavers  from  tertiary  level  education  are  much  more  likely  to  attain  professional  positions 
already  at  the  outset  of  their  careers.  In  addition,  van  der  Velden  and  Wolbers  (2000) 


establish  an  effect  of  the  strictness  of  employment  protection  on  labour  market  entrants'  job 
outcomes.  Paradoxically  at  first  glance,  the  probability  of  obtaining  initial  employment  on  a 
fixed-term  or  temporary  basis  is  higher  in  countries  with  stricter  employment  protection 
legislation.  In  fact,  this  might  indicate  a deliberate  strategy  to  flexibilise  youth  labour  markets 
so  as  to  facilitate  youth  labour  market  integration,  without  at  the  same  time  sacrificing 
protection  standards  for  the  core  work  force  (cf.  Schroder,  2000).  For  the  several  other 
institutional  indicators  as  tested  in  van  der  Velden  and  Wolbers  (2000),  results  have  not 
shown  significant  effects. 

In  addition  to  these  individual  and  institutional  factors,  the  impact  of  macrostructural  context 
factors  is  far  from  negligible.  Actually,  however,  the  role  of  aggregate  macroeconomic 
conditions  is  much  less  important  for  occupational  allocation  and  employment  outcomes  than 
for  unemployment  risks  discussed  earlier.  At  best,  macroeconomic  conditions  determine  only 
to  a small  part  the  extent  to  which  those  entering  the  labour  market  are  allocated  to  lower 
level  positions.  In  tighter  labour  markets,  young  people  are  disproportionately  allocated  into 
low-skilled  and  temporary  jobs  (Gangl,  2000b;  van  der  Velden  and  Wolbers,  2000),  and  this 
allocation  pattern  is  much  less  responsive  to  cyclical  changes  than  are  unemployment  risks. 
Whether  temporary  jobs  themselves  are,  in  turn,  more  sensitive  to  the  business  cycle  than 
permanent  jobs,  as  they  may  have  the  role  of  a buffer  to  changes  in  product  demand,  is  an 
open  question  for  future  research. 

But  what  turns  out  to  be  much  more  important  to  employment  outcomes  among  labour  market 
entrants  is  the  (changing)  balance  between  individual  qualifications  and  skill  demands  on  the 
market.  Our  analyses  clearly  show  that  net  changes  in  the  relative  balance  between  supplied 
and  demanded  skills  have  important  implications  for  employment  outcomes  among  school 
leavers.  In  general,  an  increasing  supply  of  better  qualified  market  entrants  triggers  changes 
in  allocation  patterns  at  otherwise  unchanged  market  conditions  as  better  qualified  leavers 
become  substituted  for  less  qualified  ones.  As  a consequence,  increasing  levels  of  educational 
attainment  have  diminishing  individual  absolute  and  relative  advantages  as  a by-product:  on 
average,  educational  expansion  implies  lower  occupational  status  outcomes,  higher  risks  of 
low-skilled  jobs,  and  decreasing  probabilities  of  entering  professional  positions  (Gangl, 
2000b,  2000c).  In  addition,  educational  expansion  has  also  been  accompanied  by  an 
expansion  of  part-time  employment  (van  der  Velden  and  Wolbers,  2000).  Given  the  current 
O 

ERIC 


59 


cso 


trend  of  expansion  at  the  tertiary  level,  the  triggered  adjustment  reactions  have,  of  course, 
mostly  implied  declining  occupational  returns  among  tertiary  level  leavers,  and  to  a lesser 
extent,  also  among  leavers  from  upper  secondary  education  (Gangl,  2000c).  As  with  many 
other  results  reviewed  before,  there  is  no  indication  in  our  data  that  these  processes  occur 
differently  in  different  European  countries.  If  anything,  downward  substitution  pressures  have 
even  been  somewhat  stronger  among  tertiary  level  graduates  in  Southern  European  countries, 
potentially  related  to  the  strong  ongoing  catching-up  processes  in  patterns  of  educational 
attainment. 

But  as  stated  earlier,  the  net  outcome  of  these  developments  is  dependent  on  parallel  changes 
in  the  structure  of  labour  markets.  To  the  extent  that  labour  markets  begin  to  utilise  the  higher 
level  of  supplied  skills  adequately,  an  increasing  professionalisation  of  labour  market  demand 
actually  counteracts  the  effects  of  educational  expansion  as  young  people  in  all  European 
countries  benefit  from  the  increasing  availability  of  employment  positions  appropriate  to  their 
skill  levels  (Gangl,  2000c).  There  is  strong  evidence  that  such  labour  market  developments 
actually  occurred,  although  probably  somewhat  time-lagged.  That  is,  empirically,  we  do  not 
observe  particularly  pervasive  net  changes  in  occupational  outcome  patterns  over  the  past 
decades  despite  tremendous  educational  expansion  because  labour  markets  happened  to 
generate  increasing  levels  of  demand  for  high-skill  jobs.  To  understand  if  that  correlation  was 
purely  incidental  or  whether  both  developments  have  in  fact  been  closely  interrelated  and 
potentially  intensified  each  other  would  appear  as  a pressing  task  for  future  research  - not 
least  in  order  to  have  clearer  views  on  the  policy  implications  of  even  further  educational 
expansion.  It  might  be  that  diminishing  returns  to  education  in  the  short  run  are  an  expression 
of  short-run  costs  of  adjustment  to  a modernised  economic  structure,  which  are  in  part  borne 
by  those  entering  the  market  in  a period  of  restructuring.  As  this  is  still  somewhat  tentative, 
future  research  is  clearly  needed  to  provide  answers  on  the  nature  of  such  driving  forces 
behind  changing  patterns  of  occupational  allocation. 

3.5.5  Summary 

In  sum,  the  project’s  analyses  based  on  LFS  data  have  stressed  both  the  considerable 
similarity  and  also  the  substantial  heterogeneity  in  European  transition  patterns.  There  are 
striking  differences  among  EU  member  states  in  terms  of  the  levels  and  types  of  qualifications 

ERJC 


60 


which  market  entrants  have  at  their  disposal.  There  are  important  differences  between  the 
countries  in  terms  of  the  institutional  nature  of  training  provision.  There  are  excessively  large 
differences  between  countries  in  terms  of  unemployment  risks  for  those  in  their  early  career 
stages.  And  there  are  important  differences  in  the  types  of  jobs  and  the  nature  of  employment 
contracts  attained  by  young  people.  In  fact,  the  project’s  exploratory  attempts  to  describe  the 
variety  of  European  transition  patterns  did  not  result  in  any  definite  picture  of  a European 
'map'  of  transition  experiences.  Still,  it  is  probably  fair  to  conclude  that  we  have  been  able  to 
bring  out  some  contours  more  clearly  than  can  be  done  on  the  basis  of  previous  research: 
while  the  exceptional  position  of  traditional  dual  system  countries  has  already  been  the  matter 
of  much  scholarly  debate,  notably  in  contrast  to  various  other  Northern  European  countries, 
the  particular  conditions  applying  in  most  Southern  European  countries  have  typically  gone 
unnoticed  in  comparative  research  to  date. 

Nevertheless,  as  shown  by  the  more  advanced  analyses  in  the  project,  this  heterogeneity  of 
experiences  does  not  necessarily  defy  systematic  explanation.  That  is,  the  similarities  and 
differences  in  European  transition  patterns  described  in  this  section  can  probably  be  explained 
as  arising  from  some  general  underlying  mechanisms  which  apply  to  all  countries.  In  fact,  for 
all  the  particular  outcomes  considered,  we  have  compiled  evidence  of  the  importance  of 
individual  resources,  notably  education,  institutional  factors  and  broader  socio-economic 
context  conditions  in  generating  the  observed  transition  patterns.  Typically,  cross-national 
differences  in  the  effects  of  any  such  resources  and  structural  factors  on  youth  labour  market 
integration  are  quite  small.  For  example,  while  there  is  some  evidence  for  cross-national 
differences  in  the  labour  market  value  of  (particular  types  of)  education,  the  magnitude  of 
such  effects  is  often  far  from  compelling.  Similarly,  there  is  little  evidence  that  ongoing 
labour  market  changes  affect  young  people  in  different  ways  in  different  European  countries  - 
what  does  differ  between  countries  is  the  extent  of  macroeconomic  turbulences  rather  than 
their  effects  on  transition  outcomes. 

Still,  institutional  factors  often  attain  a prominent  place  in  the  explanation  of  cross-national 
similarities  and  differences.  Country  differences  in  macro-structural  context  conditions  are 
usually  of  limited  power  in  comparative  explanations.  In  fact,  a large  part  of  cross-national 
variation  turns  out  to  be  stable  over  time  and  cyclical  economic  changes,  thus  necessitating 
institutionally-based  explanations.  Among  these,  three  particular  institutional  complexes 


figure  prominently  in  our  results:  First,  the  institutional  structure  of  education  and  training 
systems  because  this  largely  determines  the  nature  of  qualificational  resources  available  to 
market  entrants.  Countries  where  young  people  achieve  higher  levels  of  education,  as  well  as 
those  countries  operating  large  scale  systems  of  vocational  training,  provide  young  people 
with  a better  start  into  working  life.  Second,  the  institutional  labour  market  context,  which 
governs  the  transformation  of  educational  resources  into  employment  outcomes.  There  are 
two  aspects  which  have  been  addressed  more  extensively  in  project  work,  namely  the  role  of 
occupational  labour  markets  and  the  effects  of  employment  protection  legislation.  It  seems 
that  occupationalised  labour  market  contexts,  that  is,  those  labour  markets  tightly  structured 
by  occupational  boundaries  arising,  for  example,  from  the  nature  of  educational  supply  and/or 
union  action  in  recruitment  processes,  provide  some  advantages  to  young  people  as  job 
competition  relies  more  strongly  on  skills  rather  than  experience.  Hence,  the  relative 
competitiveness  of  young  people  is  increased  relative  to  systems  more  reliant  on  experience 
on  the  market. 

Finally,  there  is  the  issue  of  potential  effects  of  employment  protection  legislation,  which  is 
often  expected  to  negatively  affect  youth  labour  market  integration.  We  have  not  found  any 
evidence  which  would  support  this  assumption.  Rather,  the  evidence  seems  rather  more 
consistent  with  a view  that  in  more  tightly  regulated  systems,  the  use  of  flexibly  regulated 
forms  of  employment  contracts  (like  fixed-term  contracts  or  special  forms  of  combined  work- 
training contracts)  is  particularly  widespread  as  a regulated  means  to  foster  the  integration  of 
young  people  into  the  labour  market,  rather  than  regulation  amounting  to  a genuine 
impediment  to  integration  itself.  Countries  like  Italy  and  Greece  are  probably  among  those 
countries  where  such  an  argument  is  least  likely  to  sound  plausible,  as  tight  employment 
protection  standards  are  enforced,  but  provisions  for  flexible  contracts  to  achieve  youth 
integration  are  not  really  common,  and  both  youth  unemployment  rates  and  the  proportion  of 
first-time  job  seekers  among  the  unemployed  are  substantial.  Still,  we  have  little  direct  or 
indirect  evidence  for  a destructive  role  of  employment  protection  in  the  sense  that  employers 
appear  particularly  hesitant  to  recruit  school  leavers.  In  fact,  two  results  we  have  obtained 
might  be  taken  as  indicative  of  the  potentially  crucial  role  of  differences  in  supply-side 
behaviour  for  bringing  about  particular  transition  patterns  in  Southern  Europe:  empirically, 
what  is  specific  about  unemployment  in,  for  example,  Italy  or  Greece  is  the  extent  to  which 
leavers  from  upper  secondary,  and  even  more  so  from  tertiary,  levels  of  education  are 


affected.  This  is  very  much  in  contrast  with  what  would  have  to  be  expected  if  employment 
protection  were  the  main  problem.  In  addition,  there  is  some  evidence  that  cyclical 
fluctuations  in  youth  unemployment  rates  are  considerably  lower  in  most  Southern  European 
countries,  which  might  also  tentatively  indicate  a less  dominant  role  of  demand-side 
behaviour  in  shaping  transition  outcomes  for  young  people. 

3.6  MAIN  FINDINGS  FROM  ANALYSES  OF  THE  SCHOOL  LEAVERS'  SURVEY 
DATA 

Section  3.5  has  presented  the  main  findings  of  analyses  of  Labour  Force  Survey  data  for 
fifteen  European  countries.  In  contrast,  school  leaver  survey  data  are  available  for  only  five 
countries:  France,  Ireland,  the  Netherlands,  Scotland  and  Sweden.  While  certain  types  of 
systems  (in  particular,  dual  system  and  Mediterranean  countries)  are  excluded  from  this 
group,  important  differences  related  to  distinct  dimensions  of  educational  differentiation  and 
forms  of  labour  market  regulation  are  captured  within  the  group. 

All  of  the  systems  can  be  regarded  as  highly  standardised  but  differ  in  the  extent  and  nature  of 
differentiation  within  the  same  stage,  and  at  the  end  of  each  stage,  of  education.  The 
Netherlands  has  the  most  highly  track-differentiated  system  with  a distinction  at  both  lower 
and  upper  secondary  level  between  (different  types  of)  academic  and  vocational  courses.  In 
France,  there  is  a significant  degree  of  tracking  at  upper  secondary  level,  with  different  types 
of  lycees  and  students  studying  for  the  BEP/CAP  or  different  types  of  general  or  vocational 
Baccalaureat.  Sweden  represents  an  intermediate  case,  with  over  half  of  those  at  upper 
secondary  level  taking  vocational  programmes,  albeit  ones  with  a strong  general  component 
and  little  institutionalised  linkage  to  the  labour  market. 

Ireland  can  be  broadly  characterised  as  a 'general'  educational  system,  although  track 
differentiation  at  upper  secondary  level  has  become  increasingly  apparent  in  recent  years. 
Post-Leaving  Certificate  (PLC)  vocational  courses  are  provided  within  the  school-based 
systems  while  two  new  programme  options  (the  Vocational  and  Applied  Programmes)  have 
become  available  within  the  general  upper  secondary  examination  system.  Scotland  probably 
represents  the  clearest  example  of  an  undifferentiated  school-based  system,  albeit  with  a 
number  of  students  taking  a mix  of  academic  Highers  and  vocational  modules.  For  some 
purposes,  upper  secondary  provision  in  Scotland  can  be  seen  as  encompassing  a range  of 


differentiated  provision,  including  full-time  school,  Further  Education  (typically  vocational  or 
pre-vocational)  and  work-based  training  provision.  However,  because  of  the  nature  of  the 
sample  in  the  Scottish  school  leavers’  survey,  for  the  purposes  of  this  study  we  focus  only  on 
leavers  from  the  school-based  system,  counting  other  forms  of  upper  secondary  provision  as 
equivalent  to  early  labour  market  destinations. 

The  five  countries  also  differ  in  the  nature  of  formal  differentiation  at  the  end  of  each 
educational  stage.  Some  systems  (such  as  Ireland)  have  a highly  differentiated  grading 
structure  with  examination  candidates  awarded  grades  for  individual  subjects  which  may  be 
taken  at  a number  of  curricular  levels.  Differentiated  grading  systems  are  also  employed  in 
Scotland,  Sweden  and  the  Netherlands.  In  contrast,  systems,  such  as  France  and  the 
Netherlands,  differentiate  only  between  'passing'  and  'failing'  a particular  stage. 

In  sum,  while  school  leaver  survey  data  are  unable  to  depict  the  whole  range  of  education, 
training  and  labour  market  systems  across  Europe,  they  are  nonetheless  able  to  capture 
important  dimensions  of  institutional  variation  in  the  transition  process.  In  total,  twelve 
working  papers  were  prepared  using  the  integrated  school  leavers'  databases  (see  Table  3.5). 
These  papers  explored  a range  of  topics,  including  cross-national  variation  in  transition 
processes,  participation  in  post-school  training  and  differences  among  groups  of  young  people 
in  terms  of  gender  and  ethnicity.  The  following  sections  outline  the  main  findings  of  these 
papers  in  terms  of  (i)  educational  outcomes,  (ii)  the  relationship  between  education  and  the 
labour  market,  (iii)  post-school  training,  and  (iv)  the  social  structuration  of  transition 
processes. 


O 

ERJC 


64 


71 


Table  3,5:  Overview  of  SLS  working  papers 


Author(s) 

Title 

Topic(s) 

Grelet,  Y.,  Mansuy, 
M.,  Thomas,  G. 
(2000a) 

Transition  from  school  to  work  and  early 
labour  force  history 

Cross-national  differences  and 
similarities  in  the  nature  of 
transition  processes 

Grelet,  Y.,  Mansuy, 
M.,  Thomas,  G. 
(2000b) 

The  transition  process:  towards  exclusion 
or  financial  sufficiency,  a French-Irish 
comparison 

Prevalence  of  non-employment 
and  low  pay  over  the  first  five 
years  in  the  labour  market 

Hartkamp,  J.  and 
Rutjes,  H.  (2000a) 

A route  to  skills:  a comparative  analysis 
of  the  position  of  apprenticeship  in 
transition  systems  in  France,  Ireland,  the 
Netherlands  and  Scotland 

The  level  and  nature  of  post- 
school apprenticeship 
participation  across  countries 

Hartkamp,  J.  and 
Rutjes,  H.  (2000b) 

Apprenticeship  in  Ireland,  the 
Netherlands  and  Scotland:  comparison  of 
trends  1979-1997 

The  level  and  nature  of  post- 
school apprenticeship 
participation  across  countries 
and  over  time 

Iannelli,  C.  (2000) 

School  effects  on  youth  transitions  in 
Ireland,  Scotland  and  the  Netherlands 

Variation  between  schools 
within  countries  in  post-school 
principal  activity 

Iannelli,  C.  and  Raffe, 
D.  (2000) 

Vocational  upper-secondary  education 
and  the  transition  from  school  to  work 

Comparison  of  vocational  and 
academic  routes  in  Ireland,  the 
Netherlands,  Scotland  and 
Sweden 

Mansuy,  M.  and 
Schroder,  L.  (2000) 

Immigrant  youth  in  the  labour  market  in 
France  and  Sweden 

Educational  and  labour  market 
characteristics  of  immigrant 
youth  in  the  two  countries 

McCoy,  S.  (2000a) 

Relative  labour  market  disadvantage 
amongst  the  least  qualified  in  Ireland, 
Scotland,  the  Netherlands,  France  and 
Sweden 

Cross-national  variation  in  the 
labour  market  position  of  the 
least  qualified 

McCoy,  S.  (2000b) 

Relative  labour  market  disadvantage 
among  the  least  qualified:  Ireland,  the 
Netherlands  and  Scotland,  1979-1997 

Variation  across  countries  and 
over  time  in  the  labour  market 
position  of  the  least  qualified 

Schroder,  L.  (2000) 

The  role  of  youth  programmes  in  the 
transition  from  school  to  work 

Level  and  nature  of 
participation  in  youth 
programmes  across  countries 

Smyth,  E.  (2000a) 

Gender  differentiation  in  education  and 
transition  outcomes 

Cross-national  differences  and 
similarities  in  education  and 
transition  outcomes  among 
women  and  men 

Smyth,  E.  (2000b) 

Gender  differentiation  in  education  and 
early  labour  market  outcomes  over  time: 
a comparative  analysis 

Variation  across  countries  and 
over  time  in  education  and 
transition  outcomes  among 
women  and  men 

O 

ERIC 


65 


72 


3.6.1  Educational  outcomes 


Certain  measures  of  educational  attainment,  such  as  CASMIN  and  ISCED,  had  been 
commonly  used  in  previous  cross-national  studies.  However,  for  our  purposes  these  measures 
proved  problematic.  Firstly,  school-leavers  in  our  samples  were  not  necessarily  at  the  end  of 
their  education/training  career.  Secondly,  these  measures  often  ignored  some  of  the 
dimensions  of  education  which  are  conceptually  and  empirically  important,  at  least  in  certain 
national  contexts  (such  as  examination  grades).  For  this  reason,  we  derived  several 
dimensions  of  educational  outcomes  which  we  could  use  to  capture  the  full  complexity  of 
cross-national  variation;  these  included  age  on  leaving  school,  educational  level 
(incorporating  stage  and  qualifications  achieved),  curricular  track,  grades  received  and 
subjects/courses  taken  (see  CATEWE,  1999;  Brannen,  Smyth,  2000). 


Figure  3.9:  Cross-national  variation  in  educational  level  (1995/7) 

40 
30 
S?  20 


Dropout  Lower  level  leaving 

Source:  Calculated  from  McCoy  (2000a). 

It  was  hypothesised  that  the  institutional  nature  of  the  educational  system  would  influence  the 
proportion  of  young  people  exiting  at  different  stages  of  their  schooling.  More  specifically,  it 
was  expected  that  more  general  educational  systems  (such  as  Ireland)  would  have  a higher 
proportion  of  less  qualified  leavers  due  to  their  comparative  failure  to  retain  those  less 
academically  oriented  (McCoy,  2000a).  While  significant  cross-national  differences  were 
apparent,  this  hypothesis  was  not  wholly  confirmed.  Two  distinct  measures  of  lower 
qualifications  were  tested:  'drop-out',  exiting  the  school  system  without  any  qualifications, 
and  'lower  level  leaving',  exiting  the  school  system  without  attempting  upper  secondary 
qualifications,  a measure  that  also  included  'drop-out'.  The  nature  of  cross-national  variation 
depended  on  the  precise  measure  of  educational  level  used  (see  Figure  3.9).  Rates  of 'drop- 


■ Ireland 
H Netherlands 
B Scotland 
E Sweden 


out'  (no  qualifications)  in  the  late  1990s  were  higher  in  Scotland  than  in  Sweden,  Ireland  or 
the  Netherlands5.  If  lower  level  leaving  (no  upper  secondary  qualifications)  is  considered, 
cross-national  differences  persist  but  the  gap  in  exit  rates  between  Scotland  and  the 
Netherlands/Ireland  is  somewhat  reduced  in  magnitude  (McCoy,  2000a).  Lower  level  leaving 
remains  much  less  prevalent  in  Sweden  than  in  the  other  countries.  Achieving  an  upper 
secondary  qualification  is  highest  in  Sweden  and  Ireland  and  lowest  in  Scotland.  It  should  be 
noted  that  these  patterns  refer  to  initial  school-based  education  and  do  not  incorporate  other 
routes  to  upper  secondary  qualifications  (such  as  apprenticeships  or  further  education 
courses).  The  focus  on  school-based  educational  attainment  has  different  implications  for 
different  countries;  in  Scotland,  for  example,  many  young  people  obtain  upper  secondary 
qualifications  in  post-school  further  education  colleges  (see  Martin  and  Raffe,  1998).  Cross- 
national differences  in  educational  attainment  are  not  static  over  time,  however  (Muller  and 
Wolbers,  1999).  Among  the  SLS  countries  on  which  data  are  available,  Ireland  and  Scotland 
have  both  experienced  a substantial  growth  over  the  1980s  and  1990s  in  the  proportions 
staying  in  school  to  upper  secondary  level.  In  contrast,  educational  attainment  levels  in  the 
Netherlands  have  remained  relatively  static,  albeit  over  a shorter  time-period  (1989  to  1997) 
(McCoy,  2000b;  Smyth,  2000b). 

Young  people  in  the  countries  studied  differ  not  only  in  their  level  of  education  but  also  in  the 
type  of  education  they  receive.  Differentiation  between  'academic'  and  'vocational'  tracks  is 
clear-cut  within  the  Dutch  and  French  contexts.  In  Sweden,  vocational  specialisations  at 
upper  secondary  level  were  introduced  in  1970  with  reforms  in  the  1990s  resulting  in  a 
reduction  in  the  number  of  vocational  programmes.  In  the  Irish  and  Scottish  cases,  however, 
no  such  formal  tracking  exists  within  the  general  secondary  school  system,  although  students 
can  differ  markedly  in  the  subjects  they  take.  Furthermore,  the  Irish  system  has  seen  an 
increasing  incorporation  of  vocational  education  (most  commonly,  Post-Leaving  Certificate 
courses)  into  the  upper  secondary  level.  For  our  purposes,  this  distinction  between  the  courses 
taken  in  Ireland  and  Scotland  was  seen  as  'informal'  tracking  and  the  use  of  this  concept 
allowed  us  to  explore  whether  this  'informal'  tracking  operates  in  a similar  fashion  to  more 
formally  differentiated  tracking.  Given  the  institutional  differences  in  the  role  of  vocational 
education,  it  is  hardly  surprising  that  those  taking  vocational  tracks  represent  a larger  group  in 

5 Rates  of  drop-out  were  also  high  in  the  French  pattern.  However,  this  pattern  should  be  interpreted  with  some 
caution  due  to  the  exclusion  of  general  Baccalaureat  candidates  from  the  survey  sample.  The  French  pattern 
alters  markedly  when  this  is  taken  into  account  (see  Martin  and  Raffe,  1998). 

0 

ERIC 

n a 


61 


the  Netherlands,  Sweden  and  France  (Iannelli  and  Raffe,  2000).  Participation  in  vocational 
tracks  has  remained  fairly  stable  in  the  Netherlands  over  the  period  1989  to  1997.  However, 
there  has  been  a significant  increase  in  the  proportion  of  young  Irish  people  taking  vocational 
courses,  reflecting  the  introduction  and  expansion  of  Post-Leaving  Certificate  provision  over 
the  period  in  question6  (Smyth,  2000b). 

Within  the  countries  considered,  educational  outcomes  vary  by  gender,  ethnicity  and  socio- 
economic background;  this  variation  is  considered  in  section  3.6.4  below. 

3.6.2  Education  and  the  labour  market 

One  of  the  central  concerns  of  the  CATEWE  project  has  been  to  examine  the  relationship 
between  education  and  transition  outcomes  across  a range  of  institutional  and  labour  market 
contexts.  A number  of  measures  of  labour  market  outcomes,  such  as  the  EGP  social  class 
schema,  had  previously  been  developed  to  examine  stratification  and  mobility  processes 
among  the  adult  population.  However,  such  measures  caused  difficulties  when  applied  to  our 
particular  samples,  leading  to  a high  concentration  of  school-leavers  in  a small  number  of 
social  class  categories.  Therefore,  several  dimensions  of  labour  market  outcomes  were 
derived,  including  full/part-time  status,  the  nature  of  the  employment  contract,  social  class, 
occupational  status,  occupational  segment,  industrial  sector,  industrial  segment  and  earnings. 
The  countries  vary  markedly  along  these  dimensions  with  quite  different  industrial, 
occupational  and  earnings  structures.  A consideration  of  the  factors  underlying  these  cross- 
national differences  lies  outside  the  parameters  of  our  study.  Instead,  analyses  of  the 
integrated  SLS  datasets  have  focused  on  exploring  the  way  in  which  educational  outcomes 
help  to  shape  (variation  in)  young  people's  experiences  of  early  transition  processes. 


O 

ERJC 


6 It  is  more  difficult  to  examine  trends  over  time  in  Scotland  since  the  availability  of  data  on  the  take-up  of 
vocational  modules  varies  over  time. 


68 


7$ 


Figure  3 JO:  A summary  of  the  relationship  between  education  and  transition  outcomes 


Transition 
' outcome ' 

Educational  outcome 

Educational 

level 

Educational  type 
(vocational 
track) 

Grades 

School  variation 
(controlling  for 
composition) 

Unemployment 

- 

n.s. 

(except  Ireland  -; 

Scotland  + at 
upper  secondary 
only) 

(Ireland,  Sweden, 
Scotland) 

Significant 

(Ireland, 

Netherlands) 

Further 

education 

+ 

- 

+ 

Significant 
(Ireland,  Scotland, 
Netherlands) 

Apprenticeship 

(France7, 

Netherlands) 

(France, 

Netherlands) 

Significant 

(Ireland, 

Netherlands) 

State  training 
scheme 

(Ireland, 

Scotland) 

+ 

(Netherlands, 

Sweden) 

n.s. 

(France) 

n.s. 

Not  significant 
(Ireland, 
Netherlands, 
Scotland) 

Occupational 

status 

+ 

- 

Manual 

employment 

- 

+ 

Secondary 
sector  job 

- 

- 

Part-time  job 

(except 

Scotland) 

(Ireland, 

Netherlands) 

Earnings 

+ 

(except 

Sweden) 

+ 

(Netherlands, 

France) 

Notes:  + statistically  significant  positive  relationship 

- statistically  significant  negative  relationship 
n.s.  no  significant  relationship 

shaded  area  - relationship  not  considered  in  the  working  papers 


It  should  be  noted  that  in  France  young  people  can  either  transfer  to  apprenticeship  programmes  within  their 
initial  education  (often  at  fairly  early  ages)  or  return  to  such  programmes  after  leaving  school.  This  finding 
relates  to  post-school  participation. 


0 


69 


The  presence  in  the  integrated  databases  of  information  on  distinct  dimensions  of  education 
allowed  us  to  explore  the  relative  importance  of  particular  educational  outcomes  in  different 
national  contexts.  Analyses  centred  on  four  of  these  dimensions:  differentiation  in  terms  of 
educational  level  (stage  left  school  or  a combination  of  stage  and  qualifications  received), 
between  academic  and  vocational  tracks,  within  stages  in  terms  of  examination  grades,  and 
between  schools.  The  relationships  between  these  different  dimensions  of  educational 
background  and  a number  of  transition  'outcomes'  are  summarised  in  Figure  3.10. 

It  was  hypothesised  that  educational  outcomes  would  have  a significant  influence  on 
transition  processes  among  young  people  in  all  of  the  study  countries,  due  to  the  standardised 
nature  of  the  qualifications  systems  considered.  However,  it  was  expected  that  in  more 
'general'  systems  (such  as  Ireland  and  Scotland)  educational  level  and  grades  received  would 
assume  a more  important  role  in  shaping  transition  outcomes  while  type  of  education 
(whether  academic  or  vocational)  would  be  more  important  in  track-differentiated  systems 
like  the  Netherlands  (Hannan  et  al.,  1999). 

Educational  level 

The  stage  at  which  young  people  left  school  along  with  the  qualifications  they  achieved  were 
found  to  have  significant  influences  on  transition  outcomes  in  all  of  the  countries  considered. 
Less  qualified  leavers  are  more  likely  to  be  unemployed  than  those  with  higher  qualification 
levels  (McCoy,  2000a)  and  tend  to  have  longer  spells  of  unemployment  (Grelet  et  al.,  2000a). 
Data  from  Ireland  and  France  indicate  that  those  without  qualifications  continue  to  be  at  a 
disadvantage  in  access  to  employment,  even  five  years  after  entering  the  labour  market 
(Grelet  et  al.,  2000b).  Less  qualified  leavers  are  also  less  likely  to  secure  access  to  further 
education  than  those  with  upper  secondary  qualifications  (Iannelli  and  Raffe,  2000;  Smyth, 
2000b).  In  addition,  the  type  and  quality  of  job  are  associated  with  initial  level  of  education; 
those  with  upper  secondary  qualifications  have  access  to  better  quality  jobs  while  the  least 
qualified  are  more  likely  to  have  part-time  jobs  and  receive  low  wages  (Grelet  et  al.,  2000a; 
McCoy,  2000a).  Lower  level  leavers  tend  to  be  over-represented  in  manual  employment  and 
under-represented  in  the  routine  non-manual  or  professional  classes.  They  are  more  likely  to 
be  found  in  secondary  sector  employment  within  the  manufacturing  or  construction  sectors 
and  are  less  likely  to  secure  employment  in  the  finance,  public  administration  or  professional 
service  sectors  (McCoy,  2000a). 


While  there  are  definite  similarities  across  countries  in  the  position  of  less  qualified  young 
people,  their  relative  disadvantage  tends  to  differ  across  countries  and  in  terms  of  the  labour 
market  outcome  considered.  Unemployment  risks  are  more  strongly  differentiated  by  initial 
educational  level  in  Scotland  than  in  Ireland,  Sweden  or  the  Netherlands  (McCoy,  2000a). 
Within  France,  unemployment  risks  continue  to  be  differentiated  by  educational  level  over  the 
first  five  years  in  the  labour  market  (Grelet  et  al.,  2000b).  The  distribution  of  unemployment 
after  the  initial  period  of  labour  market  entry  also  varies  cross-nationally;  in  the  Irish  context, 
unemployment  is  concentrated  within  a small  group  who  experience  longer  term 
unemployment  while  in  the  French  context,  unemployment  is  experienced  by  a broader  group 
but  interspersed  with  periods  of  short-term  employment  and  training  programmes  (Grelet  et 
al.,  2000b).  In  terms  of  occupational  allocation,  access  to  professional  employment  is  more 
strongly  differentiated  by  education  in  Ireland  and  Scotland  than  in  the  the  Netherlands  or 
France.  Furthermore,  the  relative  disadvantage  of  lower  level  leavers  in  entry  to  secondary 
sector  jobs  is  strongest  in  the  Netherlands  (McCoy,  2000a). 

There  has  been  much  debate  about  the  growth  of  overqualification  in  the  youth  labour  market. 
While  such  studies  usually  address  labour  market  entrants  from  both  secondary  and  tertiary 
levels  (see,  for  example,  Hannan  et  al.,  1998),  it  might  be  expected  that  increasing  educational 
levels  coupled  with  growing  or  volatile  unemployment  rates  would  have  some  implications 
for  changes  in  the  returns  to  education  among  secondary  leavers.  No  consistent  picture  of 
changes  in  educational  returns  over  time  emerges;  however,  there  are  tentative  suggestions  of 
declining  returns  in  some  countries  and  for  certain  labour  market  outcomes.  Both  the 
Netherlands  and  Scotland  have  experienced  a decline  in  the  gap  between  the  most  and  the 
least  qualified  in  their  relative  unemployment  risks  over  time  (McCoy,  2000b)  with  upper 
secondary  leaving  providing  diminishing  protection  against  unemployment  in  Scotland 
(Smyth,  2000b).  There  is  also  some  evidence  of  declining  differentials  between  educational 
levels  in  occupational  status  in  the  Netherlands  and  in  secondary  sector  employment  in 
Scotland  and  Ireland  (McCoy,  2000b). 

Type  of  education 

The  type  of  education  received  (whether  academic  or  vocational)  is  considered  in  a number  of 
analyses.  It  was  hypothesised  that  the  existence  of  formal  track  differentiation  would  mean 

O 

ERJC 


71 


78 


that  type  of  education  would  have  the  strongest  effects  in  the  Dutch,  French  and,  to  a lesser 
extent,  Swedish  systems.  However,  it  was  also  recognised  that  'informal'  tracking  may  play  a 
role  in  shaping  transition  outcomes  in  Ireland  and  Scotland.  Differentiation  in  the  tracks 
young  people  take  through  the  secondary  school  system  is,  on  average,  highly  predictive  of 
their  trajectories  on  leaving  school.  Those  who  take  academic  courses  are  much  more  likely  to 
enter  further  education  than  those  who  have  taken  vocational  courses.  This  difference  is  more 
marked  in  Netherlands  and,  perhaps  surprisingly,  Ireland  than  in  Scotland8  (Iannelli  and 
Raffe,  2000).  Thus,  in  the  Irish  and  Dutch  cases,  school-based  vocational  courses  provide 
occupationally-specific  skills  and  therefore  act  as  an  alternative  to  acquiring  such  skills 
through  further  education.  Analyses  of  the  LFS  data  indicate  that  vocational  education  tends 
to  be  associated  with  a lower  unemployment  risk  (see  above).  However,  this  pattern  was  not 
apparent  in  analyses  of  the  school  leavers'  survey  data,  except  in  Ireland  where  young  people 
who  have  taken  vocational  courses  have  lower  unemployment  rates  than  other  groups 
(Iannelli  and  Raffe,  2000;  McCoy,  2000a).  It  is  likely  that  the  differences  between  the  two 
data  sources  can  be  accounted  for  by  differences  in  the  nature  of  the  samples  (apprentices 
were  an  'outcome'  for  SLS  but  'leavers'  for  LFS  purposes),  in  the  definition  of  vocational 
education  (with  a broader  definition  adopted  in  the  SLS  case)  and  the  fact  that  unemployment 
rates  in  the  highly  track-differentiated  Dutch  case  were  extremely  low  at  the  time-point 
considered. 

The  type  of  education  received  was  also  found  to  have  consequences  for  the  nature  of 
employment  achieved.  Having  taken  a vocational  track  increases  the  likelihood  of  entering 
manual  employment,  particularly  skilled  manual  jobs,  and  appears  to  provide  some  protection 
against  entry  to  secondary  sector  jobs  in  France,  Ireland,  the  Netherlands,  Scotland  and 
Sweden  (McCoy,  2000a).  Conversely,  those  who  have  taken  a vocational  track  are  much  less 
likely  than  academic  leavers  to  enter  professional  employment  or  to  find  work  in  the 
distribution,  finance  or  public  administration  sectors  (McCoy,  2000a).  It  was  hypothesised 
that  acquiring  occupationally-specific  skills  through  school-based  vocational  education  would 
also  have  a return  in  terms  of  pay  levels.  This  hypothesis  was  confirmed  with  vocational 
leavers  found  to  be  at  an  earnings  advantage  in  the  more  track-differentiated  systems  of  the 
Netherlands  and  France  (McCoy,  2000a). 


Unfortunately,  France  could  not  be  included  in  these  analyses.  As  general  Baccalaureat  leavers  were  not 
included  in  the  sample,  a full  comparison  of  academic  and  vocational  tracks  at  the  upper  secondary  level  could 
not  be  undertaken. 


72 


79 


The  distinction  between  vocational  and  academic  tracks  appears  to  be  an  important  one  in 
shaping  young  people's  early  labour  market  experiences.  In  many  ways,  informal  tracking  in 
the  Irish  context  appears  to  operate  in  a somewhat  similar  fashion  to  more  formal  tracking  in 
the  Netherlands  and  France.  This  pattern  should  be  interpreted  with  some  caution,  however, 
since  the  effect  primarily  relates  to  participation  in  Post-Leaving  Certificate  courses  which,  in 
many  ways,  are  more  advanced  in  content  as  well  as  more  occupationally  specific  than 
regular  upper  secondary  courses  (see  Iannelli  and  Raffe,  2000).  It  is  also  important  to  go 
beyond  a simple  academic/vocational  dichotomy  to  examine  the  type  of  vocational  courses 
taken  by  young  people.  One  such  approach  is  to  examine  the  gender  composition  of  different 
vocational  tracks9  (see  Smyth,  2000a;  Smyth,  2000b).  Many  of  the  consequences  of 
vocational  education  relate  primarily  to  participation  in  the  type  of  tracks  usually  dominated 
by  young  men,  which  increases  the  likelihood  of  entering  skilled  manual  employment  in  the 
manufacturing  or  construction  sectors.  However,  participants  in  mixed  or  female-dominated 
tracks  are  more  likely  to  work  in  personal  services  or  other  non-manual  employment  (Smyth, 
2000a;  2000b).  In  this  way,  educational  segregation  plays  a role  in  reproducing  gender 
segregation  within  the  labour  market,  although  segregation  is  still  apparent  among  young  men 
and  women  who  take  the  same  kinds  of  vocational  courses  (see  section  3.6.4  below). 

Examination  grades 

The  third  dimension  considered  in  analyses  of  the  SLS  databases  related  to  differentiation 
within  stages  through  examination  grades.  It  was  hypothesised  that  grades  would  have  a more 
significant  effect  on  transition  outcomes  in  more  general  education  systems  than  in  more 
track-differentiated  systems.  This  hypothesis  was  confirmed  by  the  analyses.  Among  upper 
secondary  leavers,  grades  are  associated  with  further  education  entry  in  Scotland,  Ireland  and 
Sweden.  Interestingly,  grades  are  also  associated  with  further  education  entry  in  the 
Netherlands.  The  latter  pattern  is  likely  to  relate  to  some  form  of  unmeasured  heterogeneity 
among  academic  leavers  (e.g.  greater  interest  in  further  education  among  higher-performing 
students)  since  access  to  third-level  education  is  not  usually  based  on  a numerus  clausus 
system  as  it  is  in  the  Irish  case.  Grades  also  appear  to  be  used  by  employers  in  making 
recruitment  decisions  since  higher-performing  students  have  reduced  unemployment  risks  in 

9 More  research  is  needed  on  the  relationship  between  field  of  education  and  the  nature  of  the  transition  process. 
Analysing  the  gender  mix  of  vocational  tracks  is  a way  of  assessing  the  relationship  between  educational  and 
occupational  segregation.  However,  it  would  also  be  useful  to  know  if  certain  types  of  course  content  (e.g. 


Ireland,  Sweden  and  Scotland10  (Iannelli  and  Raffe,  2000;  Iannelli,  2000).  As  might  be 
expected,  grades  are  not  significantly  associated  with  unemployment  chances  (at  least  among 
upper  secondary  academic  leavers)  in  the  Netherlands  where  type  and  level  of  education  play 
a more  important  role  (Iannelli  and  Raffe,  2000). 

School  differences 

The  role  of  school  factors  in  shaping  the  transition  process  has  been  neglected  in  transitions 
research,  even  though  school  differences  in  educational  outcomes  have  been  long  recognised. 
Iannelli.  (2000)  examined  school  differences  in  principal  activity  one  to  one  and  a half  years 
after  leaving  the  school  system  in  Ireland,  Scotland  and  the  Netherlands.  In  particular, 
significant  school  differences  were  found  in  relation  to  access  to  further  education  and 
employment.  The  basis  for  school-level  differences  was  found  to  differ  across  countries,  with 
curriculum  type  acting  as  the  main  source  of  variation  in  the  Netherlands  while  exam  grades 
and  social  mix  of  the  school  were  found  to  play  a role  in  the  Irish  and  Scottish  contexts* 11. 

3.6.3  Post-school  training 

The  discussion  in  section  3.6.2  has  focused  on  young  people's  transitions  into  further 
education  and  (un)employment.  However,  analyses  of  the  SLS  databases  have  also  yielded 
very  rich  information  on  young  people's  experiences  of  training  after  leaving  school.  This 
section  considers  two  types  of  post-school  training:  apprenticeship  and  participation  in  youth 
programmes/schemes.  The  five  countries  considered  differ  in  their  prevalence  of  these  forms 
of  training,  in  the  way  participation  in  these  forms  of  training  has  evolved  over  time  and  in 
the  way  they  relate  to  the  initial  education  system. 

Apprenticeship 

Analyses  of  the  Labour  Force  Survey  have  tended  to  treat  apprentices  as  part  of  the  initial 
education  system  (Muller  et  al.,  1999),  although  other  analyses  have  explicitly  focused  on  the 
nature  of  this  'double  status'  (Welters  and  Wolbers,  1999).  Due  to  the  nature  of  the  sample 


carpentry)  lead  to  different  transition  outcomes  in  different  countries. 

10  Information  on  grades  is  not  collected  in  the  French  survey  precisely  because  it  is  not  seen  as  a significant 
factor  in  access  to  employment. 

11  Between-school  variation  may  also  reflect  differences  in  local  labour  market  conditions  but  this  could  not  be 
examined  within  the  present  study. 


design  of  national  transition  surveys12,  apprenticeships  are  treated  as  transition  'outcomes'  for 
the  purposes  of  analysis.  The  nature  of  the  apprenticeship  system  differs  across  the  four 
countries  studied13  both  in  its  structure  and  in  participation  levels.  Among  all  school-leavers, 
apprenticeship  levels  are  highest  in  France14,  followed  by  Scotland  and  the  Netherlands,  and 
lowest  in  Ireland  (Hartkamp  and  Rutjes,  2000a).  Participation  levels  are  somewhat  more 
variable  cross-nationally  when  only  labour  market  entrants  are  considered.  In  addition,  longer 
term  trends  in  apprenticeship  participation  differ  markedly  across  countries.  In  Ireland  and 
Scotland,  male  apprenticeship  rates  contracted  in  line  with  declining  employment  during  the 
1980s  and  early  1990s,  only  recovering  in  the  mid/later  1990s,  while  Dutch  rates  declined 
substantially  during  the  mid-1990s  (see  Figure  3.11).  Across  all  four  countries,  participation 
in  apprenticeships  is  highly  gendered,  with  significantly  higher  rates  evident  among  young 
men  (see  section  3.6.4  below). 


Figure  3.11:  Proportion  of  apprentices  among  labour  market  entrants 


Across  the  four  countries,  the  apprenticeship  system  plays  a distinctive  role  in  relation  to  the 
initial  education  system  and  to  the  labour  market.  Apprenticeships  form  an  alternative  to 


12  The  exception  to  this  is  France  where  there  are  two  groups  of  apprentices:  one  part  of  the  initial  education 
system,  the  other  those  who  re-enter  apprenticeships  after  a period  outside  the  educational  system. 

For  the  time-period  to  which  data  relate,  no  formal  apprenticeship  system  existed  in  Sweden. 

The  French  pattern  may,  at  least  in  part,  reflect  the  nature  of  the  sample  since  general  Baccalaureat  leavers  are 
not  included. 


school-based  vocational  education  as  a means  of  acquiring  occupationally-specific  skills  in 
France  and  the  Netherlands  (Hartkamp  and  Rutjes,  2000a);  those  who  have  taken  upper 
secondary  vocational  tracks  are  much  less  likely  to  enter  apprenticeships  in  these  countries 
(Iannelli  and  Raffe,  2000;  Smyth,  2000a).  In  addition,  apprentices  in  these  systems  tend  to 
have  lower  educational  attainment  levels  than  those  in  employment.  In  contrast,  in  Ireland  and 
Scotland  apprenticeship  operates  as  a type  of  post-school  vocational  training  and  apprentices 
tend  to  resemble  the  employed  in  their  educational  profile  (Hartkamp  and  Rutjes,  2000a).  The 
educational  profile  of  apprentices  has  changed  considerably  over  time,  however.  Over  time, 
the  relative  advantage  of  upper  secondary  leavers  in  securing  access  to  apprenticeships 
(compared  with  becoming  unemployed)  has  declined  over  time.  The  reverse  is  true  in  Ireland 
where  the  relative  representation  of  upper  secondary  leavers  among  apprentices  increases 
over  time  (Smyth,  2000b).  There  are  significant  differences  across  countries  and  over  time  in 
the  occupational  positions  of  apprentices,  with  skilled  manual  work  dominating 
apprenticeships  to  a particularly  marked  degree  in  Ireland.  Apprentices  enter  a broader  range 
of  occupations  in  the  Netherlands  and  France.  In  Scotland,  the  occupational  range  of 
apprenticeship  has  broadened  somewhat  over  time  but  remains  narrower  than  the  range  in  the 
Netherlands  and  France  (Hartkamp  and  Rutjes  2000a;  2000b). 


Youth  programmes/schemes 

As  with  apprenticeship,  the  prevalence  of  youth  programmes  varies  cross-nationally  and  over 
time.  In  the  late  1990s,  those  on  youth  programmes  made  up  a higher  proportion  of  school- 
leavers  in  Scotland,  France  and  Sweden  than  in  Ireland  or  the  Netherlands  (Schroder,  2000). 
In  Ireland,  Scotland  and  Sweden,  youth  programmes  emerged  as  a response  to  increasing 
unemployment  among  young  people  in  the  1970s  and  1980s.  However,  overall  levels  of 
provision  in  Scotland  significantly  exceeded  those  in  Ireland  to  the  point  where  many  of  the 
previously  traditional  employment  opportunities  for  school  leavers  were  replaced  with  youth 
programme  places,  a pattern  that  also  became  evident  in  the  French  situation.  In  the 
Netherlands,  schemes  were  not  introduced  until  later  than  in  the  other  countries  and  overall 
levels  of  participation  among  young  people  in  schemes  have  remained  low. 


O 

ERIC 


It  was  hypothesised  that  the  nature  of  labour  market  regulation,  in  combination  with  the 
prevalence  of  linkages  from  the  education  system  to  the  labour  market,  would  affect  both  the 
level  and  nature  of  youth  programme  provision.  Thus,  levels  of  provision  would  be  higher 
where  strictly  regulated  labour  markets  are  combined  with  weak  linkages  from  the  education 


76 


83 


system  to  the  labour  market.  The  hypothesised  relationship  between  regulatory  frameworks, 
the  education  system  and  youth  programme  provision  was  only  partially  confirmed  by  the 
data  (Schroder,  2000).  Participants  did,  indeed,  form  a high  proportion  of 'at  risk'  youth  (the 
unemployed  plus  those  on  programmes)  in  Sweden  and  France,  two  of  the  most  regulated 
labour  markets;  however,  participation  levels  are  also  high  in  Scotland  where  labour  markets 
are  more  'flexible'.  This  pattern  may  be  explained  by  the  emergence  of  youth  programmes  as 
another  form  of  post-school  vocational  training  in  the  Scottish  context. 

In  the  more  regulated  systems  of  France,  Sweden  and  the  Netherlands,  the  role  of 
programmes  appears  to  be  to  provide  school-leavers  with  work  experience  while  schemes 
play  a role  in  providing  general  or  specific  training  in  the  other  countries  considered.  The 
educational  profile  of  programme  participants  varies  across  countries.  There  is  a negative 
selection  among  those  'at  risk'  in  Scotland  and  Ireland,  that  is,  programmes  tend  to  be  targeted 
on  less  qualified  young  people.  In  contrast,  selection  is  positive  in  the  Netherlands,  with 
higher  participation  among  upper  secondary  academic  leavers15,  and  Sweden,  with  higher 
participation  among  all  upper  secondary  leavers.  In  France,  participation  is  not  significantly 
related  to  prior  education  (Schroder,  2000). 

3.6.4  The  social  structuration  of  transition  processes 

A number  of  research  studies  have  indicated  that  transition  processes  have  become 
'individualised'  over  time  with  patterns  becoming  more  differentiated  in  terms  of  the  kinds  of 
pathways  taken  by  young  people  and  less  differentiated  in  terms  of  structural  characteristics, 
such  as  gender  and  social  class  (see,  for  example,  Roberts  et  al.,  1994.).  The  potential  for 
continuing  differentiation  between  groups  of  young  people  was  addressed  by  SLS  papers  in 
relation  to  three  dimensions:  gender,  social  class  and  ethnicity/national  origin.  The  extent  to 
which  these  issues  could  be  explored  varied  across  countries  as  not  all  national  transition 
surveys  contained  information  on  these  dimensions.  However,  analyses  do  highlight  the 
importance  of  considering  these  sources  of  potential  differentiation  in  transition  processes. 


15  This  may,  in  part,  reflect  an  age  effect  since  age  is  used  as  a criterion  for  accessing  many  youth  programmes 
in  the  Netherlands. 

O 

ERIC 

4'  C A 


Gender  • 

Educational  and  transition  outcomes  differ  by  gender  in  all  five  countries  but  the  nature  of 
these  differences  varies  cross-nationally.  Currently,  young  women  are  more  likely  to  attain 
upper  secondary  qualifications  than  young  men  in  all  five  of  the  countries  studied. 
Furthermore,  within  the  same  examination  level,  they  tend  to  outperform  their  male 
counterparts,  with  the  exception  of  the  Dutch  situation  (Smyth,  2000a).  Young  men  are  more 
likely  to  take  vocational  tracks  than  young  women  in  all  countries,  except  Scotland.  However, 
the  gender  differences  are  somewhat  reduced  when  only  upper  secondary  leavers  are 
considered.  When  a longer  term  perspective  is  taken,  Ireland,  Scotland  and  the  Netherlands 
exhibit  quite  different  trends  in  gender  differentiation  in  educational  attainment  (see  Figure 
3.12).  Scotland  shows  a widening  gender  gap  (in  favour  of  females)  in  educational 
attainment,  Ireland  shows  a slight  narrowing  of  the  gender  gap  (with  males  catching  up 
somewhat)  while  the  Netherlands  shows  relative  stability  in  the  pattern  of  gender  differences, 
albeit  over  a shorter  time  period  (Smyth,  2000b). 


Figure  3.12:  Gender  differences  in  upper  secondary  completion  over  time 


Irish  males 

Irish  females 

— * — Scottish  males 
— » — Scottish  females 
11  Dutch  males 

1 Dutch  females 


It  might  be  expected  that  higher  educational  attainment  among  young  women  would  translate 
into  greater  labour  market  advantage.  Transition  outcomes  are,  indeed,  found  to  differ  by 
gender  but  differently  across  countries.  Among  labour  market  entrants,  young  women  are 
more  likely  to  enter  employment  (relative  to  unemployment)  than  their  male  counterparts, 

O 

ERIC 


78 


n pm 


even  controlling  for  their  higher  educational  levels;  this  pattern  has  been  remarkably 
consistent  over  time  (Smyth,  2000b).  Across  all  countries,  young  women  are  much  less  likely 
to  enter  apprenticeships  than  young  men,  with  even  more  marked  gender  differences  apparent 
in  the  Irish  context;  this  pattern  would  appear  to  relate  to  cross-national  differences  in  the 
types  of  sectors  which  tend  to  employ  apprentices  (Smyth,  2000a;  Hartkamp  and  Rutjes, 
2000a).  Contrary  to  some  interpretations  of  the  individualisation  hypothesis,  there  is  no 
evidence  of  a net  reduction  in  gender  differences  in  transition  outcomes  over  time  (Smyth, 
2000b). 

Higher  educational  levels  among  young  women  do  not  consistently  translate  into  labour 
market  advantage.  Firstly,  young  women  are  found  to  be  more  highly  concentrated  in  part- 
time  work  than  their  male  counterparts  and  this  concentration  has  increased  over  time16 
(Smyth,  2000b).  Secondly,  in  Ireland,  the  Netherlands,  France  and  Sweden,  young  women 
receive  lower  hourly  earnings  than  young  men,  all  else  being  equal.  This  pattern  is  not 
attributable  to  differences  in  occupational  status  or  training  situation  (Smyth,  2000a). 

Across  all  countries  there  is  marked  segregation  in  occupational  and  industrial  allocation  by 
gender.  It  was  hypothesised  that  educational  segregation  into  vocational  tracks  would  result  in 
higher  levels  of  labour  market  segregation  by  gender  in  the  Netherlands  than  in  the  other 
countries.  In  terms  of  industrial  and  occupational  distribution,  this  is  tentatively  confirmed. 
However,  industrial  and  occupational  outcomes  differ  by  gender  even  among  those  who  have 
taken  similar  vocational  tracks,  indicating  that  labour  market  segregation  is  at  best  only 
partially  mediated  by  the  distribution  of  young  men  and  women  across  different  vocational 
tracks  (Smyth,  2000a;  2000b).  It  is  also  worth  noting  that  educational  qualifications  do  not 
always  have  the  same  returns  for  males  and  females.  In  particular,  taking  a vocational  track 
appears  to  have  quite  different  influences  on  early  labour  market  outcomes  for  males  and 
females  (Smyth,  2000a;  2000b).  Taking  a female-dominated  track,  for  example,  is  associated 
with  higher  returns  in  terms  of  occupational  status  for  males  than  for  females. 


16  The  extent  to  which  this  pattern  reflects  constraint  or  choice  cannot  be  systematically  explored  using  the  SLS 
datasets.  However,  separate  analyses  indicate  that  those  in  part-time  jobs  in  Ireland  and  the  Netherlands  are 
much  more  likely  than  their  full-time  counterparts  to  be  actively  looking  for  another  job  and  this  pattern  is 
evident  for  both  males  and  females. 

O 

ERJC 


79 


8fi 


Social  class 

The  potential  impact  of  family  background  factors  on  transition  processes  was  considered  in 
relation  to  three  countries:  Ireland,  Scotland  and  Sweden17.  In  Ireland,  Scotland  and  Sweden, 
those  from  professional  or  other  white-collar  backgrounds  are  more  likely  to  attain  upper 
secondary  qualifications,  especially  academic  qualifications,  than  those  from  manual  class 
backgrounds  (Iannelli  and  Raffe,  2000).  In  keeping  with  previous  research  (Shavit  and 
Blossfeld,  1993),  the  nature  of  social  class  differences  in  educational  outcomes  tends  to  be 
similar  across  countries  and  in  neither  Scotland  or  Ireland  is  there  any  evidence  of  a decline  in 
the  strength  of  the  relationship  between  family  background  and  educational  attainment 
(Smyth,  2000b).  Indeed,  there  is  some  evidence  that  having  a parent  in  paid  employment  has 
become  a more  significant  factor  over  time. 

Given  the  relationship  between  socio-economic  background  and  educational  outcomes, 
family  background  will  have  an  indirect  effect  on  transition  outcomes  among  young  people. 
However,  a direct  effect  is  also  evident.  Entry  to  further  education  is  sharply  differentiated  by 
social  class  in  Ireland  and  Scotland  (Smyth,  2000b),  even  controlling  for  examination  grades 
(Iannelli,  2000).  Social  class  background  is  also  associated  with  employment  chances  and 
quality  of  jobs.  Those  from  professional  class  backgrounds  have  higher  employment  chances, 
higher  rates  of  entry  to  the  finance,  public  administration  and  service  sectors  and  higher  rates 
of  entry  to  professional  and  petty  bourgeois  employment  than  those  from  semi/unskilled 
manual  backgrounds  (Smyth,  2000b).  The  latter  findings  should  be  interpreted  with  some 
caution;  further  analysis  is  needed  to  explore  whether  these  patterns  hold  when  more  detailed 
measures  of  educational  attainment  (such  as  exam  grades)  are  taken  into  account. 

National  origin 

Ethnicity/national  origin  has  been  a neglected  area  in  comparative  transitions  research. 
However,  the  availability  of  information  on  national  origin  in  the  French  and  Swedish  surveys 
meant  that  this  issue  could  be  explored  using  the  longitudinal  database  (Mansuy,  Schroder, 
2000). 


In  both  France  and  Sweden,  those  bom  abroad  and  those  with  two  parents  bom  abroad  are 
more  likely  than  native-born  youth  to  leave  school  on  completion  of  compulsory  education. 


17  No  information  was  available  on  parental  characteristics  in  the  Netherlands  while  the  measures  of  parental 


erIc 


80 


87 


This  pattern  persists,  even  when  parental  education  and  employment  are  taken  into  account. 
Foreign-bom  youth  appear  to  be  at  a greater  disadvantage  in  education  in  France  than  in 
Sweden,  reflecting  the  selectivity  of  the  French  educational  system.  If  immigrant  young 
people  in  France  and  Sweden  do  stay  in  school,  they  are  less  likely  to  take  vocational  routes 
than  their  native-born  counterparts  which  is  likely  to  have  consequences  for  their  subsequent 
labour  market  position  (see  above). 

In  both  France  and  Sweden,  those  bom  abroad  are  more  likely  to  be  unemployed  and  less 
likely  to  be  employed  than  native-born  youth  five  years  after  leaving  school.  This  higher 
unemployment  risk  persists  even  when  educational  and  family  background  are  taken  into 
account.  Controlling  for  education,  young  people  with  immigrant  parents  are  at  a higher  risk 
of  unemployment  in  Sweden  than  in  France.  This  pattern  appears  to  reflect  the  greater 
signalling  power  of  education  in  France. 

In  both  countries,  some  ethnic  groups  are  believed  to  be  particularly  exposed  to 
discrimination  on  the  labour  market.  In  France,  the  group  experiencing  greatest  disadvantage 
are  young  people  of  North  African  origin  while  in  the  Swedish  case  North  African 
immigration  is  comparatively  rare  with  greatest  disadvantage  evident  among  those  of  Asian 
(frequently  Turkish),  Latin  American  and  other  African  origins.  In  France,  youth  of  North 
African  origin  enter  the  labour  market  with  a lower  educational  level  than  other  groups.  In 
both  countries,  young  people  belonging  to  these  'disadvantaged'  groups  have  higher 
unemployment  risks  than  other  young  people.  These  higher  unemployment  risks  remain  when 
variables  controlling  for  the  individual's  education  and  social  background  are  taken  into 
account. 

In  summary,  analyses  of  the  integrated  SLS  database(s)  indicate  that  social  class,  gender  and 
national  origin  continue  to  play  an  important  role  in  structuring  transition  processes  among 
young  people  and  show  no  evidence  of  a decline  in  the  significance  of  gender  and  social  class 
as  sources  of  variation. 


occupation  used  in  France  were  not  comparable  with  those  in  the  other  countries. 


O 

ERIC 


si  88 


3.6.5  Summary 


In  summary,  analyses  of  the  integrated  SLS  database(s)  yield  significant  insights  into  our 
understanding  of  institutional  variation  in  transition  processes  among  young  people.  Firstly, 
they  allow  us  to  explore  the  pathways  taken  by  young  people  in  different  education,  training 
and  labour  market  contexts.  Secondly,  they  highlight  potential  sources  of  heterogeneity 
among  countries  that  are  often  considered  similar  in  other  respects.  Thirdly,  they  indicate  the 
importance  of  adopting  a multidimensional  approach  to  examining  both  educational  and 
labour  market  outcomes.  The  pattern  of  cross-national  variation  can  differ  substantially 
depending  on  which  educational  or  labour  market  outcome  is  considered.  Fourthly,  they  allow 
us  to  explore  differentiation  among  groups  of  young  people  in  terms  of  their  gender,  socio- 
economic background  and  ethnicity.  The  analyses  not  only  highlight  substantive  issues 
relating  to  institutional  variation  in  young  people's  experience  of  the  transition  process  but 
they  also  contribute  to  our  understanding  of  the  requirements  for  systematic  comparative 
research.  The  following  section  highlights  some  of  the  lessons  to  be  learned  from  the 
CATEWE  experience. 

3.7  NEW  DATA  COLLECTION 

3.7.1  Developing  recommendations  for  future  data  collection 

The  weaknesses  of  SLS  data,  discussed  in  the  previous  section,  include  the  limits  to  their 
comparability  arising  from  differences  in  the  survey  designs,  data  coverage  and  data 
definitions  of  the  school  leaver  surveys.  We  had  expected  to  encounter  such  problems,  and  in 
the  CATEWE  project  proposal  we  promised  to  use  the  experience  of  constructing  and  using 
the  SLS  datasets  to  make  proposals  for  ‘the  harmonisation  of  future  school  leavers'  surveys’ 
and  for  ‘the  design  of  a European-wide  school  leavers'  survey’.  The  reports  which  presented 
the  first  analyses  of  LFS  and  SLS  data  included  preliminary  assessments  of  comparability  and 
other  issues  arising  from  both  data  sources  (Muller  et  al.,  1999;  Hannan  et  al.,  2000).  The 
project  made  presentations  to  the  September  1999  workshop  of  the  European  Research 
Network  on  Transitions  in  Youth  which  discussed  the  strengths  and  weaknesses  of  SLS  data 
for  comparative  research  (Biggart  and  Raffe,  2000;  Smyth  and  McCoy,  2000).  On  behalf  of 
the  project,  the  University  of  Edinburgh  obtained  funding  from  the  Accompanying  Measures 

er|c 


82 


89 


programme  for  an  International  Workshop  on  Comparative  Data  on  Education-to-Work 
Transitions,  held  at  the  OECD  in  Paris  in  June  2000  in  association  with  Network  B (on 
transitions)  of  the  OECD’s  Educational  Indicators  Project.  The  workshop  was  attended  by 
policy-makers,  statisticians  and  researchers  from  eighteen  countries  as  well  as  from  the 
European  Commission,  EUROSTAT,  CEDEFOP  and  the  OECD  itself.  It  received  six 
presentations  from  members  of  the  CATEWE  project,  as  well  as  presentations  from  other 
European  researchers,  EUROSTAT,  the  European  Commission  (DG  Education  and  Culture), 
representatives  of  national  governments  and  statistical  offices,  participants  in  Network  B,  and 
members  of  the  OECD  associated  with  its  OECD’s  Thematic  Review  of  the  Transition  from 
Initial  Education  to  Working  Life  and  with  the  proposed  Longitudinal  Option  to  its  2003 
PISA  survey.  A separate  report  on  the  workshop  is  in  preparation.  One  of  the  CATEWE 
papers  to  the  workshop  was  an  initial  draft  of  the  CATEWE  project's  recommendations  on 
strategies  for  cross-national  data.  This  draft  was  substantially  revised  after  the  workshop  and 
presented  to  the  September  2000  workshop  of  the  European  Research  Network  on  Transitions 
in  Youth.  After  further  revision  it  was  submitted  to  the  Commission  (Raffe,  2000). 

Drawing  on  its  own  experience,  and  the  various  discussions  described  above,  the  project 
defined  a set  of  requirements  for  a data  and  indicator  system  on  education-to-work  transitions. 
Such  a system  should 

• provide  regular  data,  that  are  comparable  over  time,  as  a basis  for  measuring  trends  and 
for  analysing  the  impact  of  changes  in  policy  and  in  institutions; 

• cover  all  stages  of  the  transition  process,  including  transitions  within  the  upper  stages  of 
education  and  training  as  well  those  during  the  early  years  in  the  labour  market  and 
through  intermediate  or  dual  statuses;  and  it  should  allow  for  ‘reverse’  transitions  from 
work  to  education  as  well  as  vice  versa ; 

• provide  longitudinal  (flow)  data  that  track  individuals  through  all  these  transitions,  in 
order  to  identify  individual  itineraries  and  in  order  to  analyse  the  determinants  of 
successful  transitions; 

• include  subjective  data  collected  before  key  decisions  on  transition  are  made; 

• provide  data  on  the  processes  of  transition  as  well  as  the  starting  points  and  outcomes; 

• provide  data  on  multiple  outcomes  of  transition,  including  learning  outcomes  as  well  as  a 
variety  of  labour-market  outcomes; 


• provide  data  on  key  family  and  household  transitions  (leaving  the  parental  home,  forming 
a stable  partnership,  becoming  a parent); 

• include  the  necessary  data  (gender,  social  background,  ethnicity/nationality)  for  the 
measurement  of  equity  issues; 

• cover  a full  cross-section  of  young  people,  so  that  inequalities  can  be  measured  against  the 
full  cohort,  and  as  a basis  for  a system-wide  perspective; 

• be  sensitive  to  national  differences  in  the  institutions  and  processes  of  transition; 

• provide  data  that  are  formally  comparable  across  countries,  in  the  sense  that  populations, 
variables  and  classifications  are  based  on  the  same  formal  definitions; 

• provide  data  that  are  substantively  comparable  across  countries:  this  means,  among  other 
things,  that  comparisons  should  not  all  be  related  to  a single  transition  event  which  may 
not  have  the  same  significance  in  each  country;  and 

• include  information  on  national  contexts  (economic  conditions,  labour  markets,  education 
and  training  systems,  government  structures,  etc)  with  which  to  interpret  these 
comparisons. 

When  the  current  main  data  sources  on  transition  (SLS,  LFS  and  administrative  data)  are 
evaluated  against  these  criteria,  not  surprisingly  none  fully  satisfies  all  the  requirements. 
Moreover,  none  can  easily  be  adapted  or  modified  to  do  so.  In  particular,  there  is  limited 
scope  for  harmonising  national  school  leaver  surveys,  which  differ  widely  in  design  and 
content,  and  which  serve  distinct  national  purposes.  The  project  outlined  a strategy  for  ‘partial 
harmonisation’  of  national  surveys,  based  on  a set  of  criteria  to  which  they  should  be 
encouraged  to  converge  when  this  does  not  conflict  with  national  priorities. 

The  project’s  main  proposal  for  future  data-collection  strategies  is  that  they  should  be  based 
on  the  complementarity  of  different  data  sources.  It  is  not  necessary  for  each  data  source 
individually  to  satisfy  all  the  criteria  listed  above,  but  collectively  they  should  do  so.  The 
project  proposes  that  a new  European- wide  survey  should  be  designed  to  fill  the  main  gaps  in 
existing  data  sources.  These  gaps  are: 

• the  collection  of  ‘equity’  variables  (gender,  social  background,  ethnicity /nationality) 
on  a consistent  basis: 

• data  on  itineraries  within  the  education  and  training  system; 

• subjective  data  collected  prior  to  key  decision  points;  and 


• ‘substantive  comparability’  between  countries:  that  is,  comparability  not  based  on  a 
single  transition  event.  The  typical  sequence  and  nature  of  transitions  varies  across 
countries,  and  no  single  transition  event  can  have  exactly  the  same  significance  within 
this  sequence  in  different  countries. 

The  project  considered  three  possible  designs  for  a cross-national  survey:  a transition  survey 
(based  on  a single  transition  such  as  leaving  secondary  school),  a prospective  age  cohort 
survey  and  a retrospective  age  cohort  survey.  A cross-national  transition  survey  - with  a 
design  similar  to  the  school  leavers'  surveys  in  the  CATEWE  project  - could  not  easily  fill  the 
gaps  listed  above.  Transition  surveys  tend  not  to  collect  details  of  transitions  within  the 
education  system;  they  can  only  collect  subjective  data  on  reasons  for  transitions 
retrospectively  (and  therefore  inadequately);  and  because  they  are  based  on  a single  transition 
they  afford  limited  substantive  comparability.  The  CATEWE  experience  illustrated  this 
problem:  the  definition  of  ‘leaving  secondary  school’  embodied  in  the  surveys  varied  across 
the  five  SLS  countries.  A cross-national  transition  survey  would  not  therefore  fill  the  gaps  in 
existing  data  sources.  The  project  recommended  a prospective  age  cohort  survey,  which 
might  be  replaced  by,  or  amalgamated  with,  the  proposed  longitudinal  survey  to  follow  up  the 
PISA  survey  of  15  year-olds  in  2003.  Our  recommendations  are  discussed  in  chapter  4 below. 

3.7.2  Support  for  new  surveys 

A further  methodological  aim  of  the  project  was  to  provide  scientific  support  for  new  national 
transition  surveys.  Two  such  surveys  were  associated  with  the  project,  with  the  research 
teams  responsible  being  engaged  as  sub-contractors. 

In  Belgium,  SONAR  - an  interdisciplinary  research-team  with  links  to  four  Flemish 
universities  and  formed  on  the  initiative  of  the  Policy  Co-ordination  Unit  of  the  Department 
of  Education  of  the  Flemish  Community  - surveyed  a representative  sample  of  3,000  Flemish 
23  year  olds  (bom  in  1976),  sampled  from  the  national  register.  Face-to-face  interviews  took 
place  between  October  1999  and  February  2000.  The  data  collected  included  details  of  young 
people’s  social  background,  secondary  and  tertiary  education  and  experiences  in  the  labour 
market.  A monthly  calendar  recorded  jobs  and  unemployment  spells  from  the  end  of 
compulsory  education  at  18  years.  It  is  planned  to  survey  a fresh  cohort  of  23  year  olds  in 
2001  and  to  contact  the  1976  cohort  again  in  2002  (at  age  26). 

O 

ERIC 


J 


85 


Association  with  CATEWE  was  instrumental  in  developing  the  SONAR  survey  in  several 

ways. 

• Even  if  it  was  clear  from  the  start  that  the  institutional  peculiarities  of  the  Flemish 
educational  system,  most  notably  the  fact  that  compulsory  schooling  age  is  set  at  18, 
prevented  a simple  replication  of  a framework  as  used  in  one  of  the  existing  national 
surveys,  the  experience  of  the  CATEWE  components  allowed  SONAR  to  decide  about 
the  format  the  Flemish  survey  on  the  basis  of  detailed  information  and  comparison  with 
regard  to  sampling  population,  interviewing  methods,  types  of  questionnaires  or  lists  of 
variables. 

• Participating  in  the  meetings  of  the  CATEWE  team  was  important  in  this  respect.  Detailed 
discussions  about  the  construction  of  internationally  comparative  databases  on  the  basis  of 
existing  surveys  gave  SONAR  much  insight  in  the  way  questions  could  (or  ought)  to  be 
phrased  in  order  to  make  variable  construction  easier  (and  to  allow  for  eventual 
international  comparability).  Early  in  the  process  of  setting  up  the  SONAR  framework, 
members  of  the  CATEWE  team  were  invited  to  a workshop  in  Brussels  to  address 
researchers,  representatives  of  the  Flemish  Department  of  Education  and  representatives 
of  the  funding  agencies  to  share  their  experience  and  to  advise  on  possible  ways  to  set  up 
a school-leaver  or  age-cohort  survey. 

• Its  association  with  CATEWE  also  allowed  SONAR  to  link  to  the  wider  European 
Research  Network  on  Transitions  in  Youth.  SONAR  hosted  the  7th  Annual  Workshop  of 
T1Y  (Antwerp,  7-10  September  2000).  First  results  of  the  Flemish  survey  were  presented 
at  the  final  day  of  this  workshop  with  members  of  the  CATEWE  project  acting  as  chair 
and  discussant. 

• Given  the  timing  of  the  SONAR  survey  and  the  differences  in  design,  the  Flemish  survey 
being  neither  a one-year  leaver  survey  nor  a five-year  follow  up,  it  has  not  yet  been 
possible  to  construct  a Flemish  part  to  add  to  the  CATEWE  database.  However,  once  the 
integration  of  the  data-sets,  generated  on  the  basis  of  both  parts  of  the  interview  (calendar 


86 


and  questionnaire),  is  completed,  SONAR  plans  to  construct  a database  based  as  closely 
as  possible  on  the  CATEWE  variable  lists. 

In  Portugal,  the  CATEWE  project  was  associated  with  the  Graduates’  Integration  Pathways 
Observation  System  (ODES).  This  project  was  run  jointly  by  four  agencies:  two  within  the 
Ministry  of  Labour  and  Solidarity  (INOFOR  - Institute  for  Innovation  in  Training;  DETEFP 
- Department  of  Statistics  for  Labour,  Employment  and  Vocational  Training),  and  two  within 
the  Ministry  of  Education  (DGESup  - General  Directorate  of  Higher  Education;  DAPP  - 
Department  of  Forecasting  and  Planning). 

In  the  initial  phase  of  the  project,  a telephone  survey  was  carried  out  among  graduates  who 
had  completed  training  courses  at  Polytechnics  and  Universities  (both  state  and  private)  in 
1993/94,  in  one  of  the  following  six  areas:  Economics,  Management,  Engineering, 
Accountancy,  Pre-School  and  Primary  School  Teaching. 

In  Portugal,  Higher  Education  institutions  inform  the  Ministry  of  Education  (General 
Directorate  of  Higher  Education)  how  many  students  complete  their  respective  courses  each 
year.  Having  completed  their  training  course  in  the  1993/94  academic  year  was  one  of  the 
criteria  for  graduates  to  be  included  in  the  survey.  According  to  official  figures,  a total  of 
10,040  students  graduated  in  1993/94.  After  updating  telephone  numbers  and  addresses, 
7,680  graduates  were  found  to  be  contactable.  Of  these  7,680,  5,288  actually  took  part  in  the 
survey,  representing  53  per  cent  of  those  who  graduated  in  1993/94. 

The  survey  was  undertaken  on  the  basis  of  an  analytical  model  encompassing  what  were 
regarded  as  four  key  research  areas  in  a study  of  the  socio-vocational  integration  pathways  of 
young  people  in  general,  and  of  graduates  in  particular:  their  social  and  educational 
background,  professional  trajectory,  and  their  views  on,  and  expectations  for,  their 
educational  and  vocational  pathways. 

The  INOFOR  team  attended  nearly  all  the  CATEWE  project’s  meetings  and  hosted  one  of 
them  in  Lisbon  in  November  1998.  Timing  differences  meant  that  the  results  of  the  Graduates 
Survey  could  not  be  incorporated  in  the  School  Leavers'  Survey  group’s  comparative 
database.  On  balance,  however,  participation  within  the  CATEWE  project  was  positive,  in 
two  ways  particularly: 


1.  The  debate  that  ensued  at  the  meetings  resulted  in  both  a better  understanding  of  each 
country’s  education  system  and  employment  market  specificities,  and  a sharing  of 
information  about  difficulties  encountered  in  building  variables  that,  while  suitable  for 
international  comparison,  would  still  reflect  the  particularities  of  the  different  national 
realities. 

2.  Taking  part  in  the  CATEWE  project  also  led  to  interaction  with  the  European  Research 
Network  on  Transitions  in  Youth,  the  next  meeting  of  which  will  be  hosted  by  INOFOR 
in  Lisbon. 


This  chapter  has  outlined  the  main  results  of  the  CATEWE  project.  The  implications  of  these 
findings  for  policy  and  future  research  are  discussed  in  the  following  chapter. 


CHAPTER  4:  CONCLUSIONS  AND  POLICY  RECOMMENDATIONS 
4.1  INTRODUCTION 


This  chapter  outlines  the  main  conclusions  and  related  policy  recommendations 
arising  from  the  CATEWE  project.  The  project  set  out  to  address  the  deficit  in 
existing  transitions  research  by  developing  a more  adequate  framework  for  examining 
the  relationship  between  education,  training  and  labour  market  contexts  in  different 
national  contexts,  and  applying  this  framework  to  empirical  studies  of  transition 
processes  in  a range  of  European  countries.  The  main  research  question  underlying 
the  project  can  be  seen  as:  "How  do  national  systems  shape  transition  processes  and 
outcomes?".  In  particular,  we  sought  to  identify: 

1.  The  nature  of  variation  (and  similarity)  in  education  and  training  systems  across 
Europe; 

2.  The  extent  of  inequalities  in  educational  outcomes  by  gender,  social  class  and 
ethnicity,  and  the  way  in  which  these  differences  may  vary  across  systems; 

3.  The  nature  of  transition  processes,  and  their  variation  across  national  systems; 

4.  The  relationship  between  (different  kinds  of)  educational  outcomes  and  transition 
processes  and  outcomes;  and, 

5.  The  extent  of  inequalities  in  transition  outcomes  (by  gender,  social  class  and 
ethnicity),  and  the  way  in  which  these  differences  may  vary  across  systems. 

The  remainder  of  this  chapter  summarises  our  conclusions  in  relation  to  these  issues 
and  identifies  the  ways  in  which  the  project  has  contributed  to  our  understanding  of 
transition  processes  across  Europe. 


Section  4.2  reviews  the  nature  of  transition  processes  and  outcomes,  highlighting  the 
distinctive  features  of  the  transition  from  education  to  work  in  Europe  at  the  turn  of 
the  millennium,  the  way  in  which  transitions  are  changing  and  the  way  in  which  they 
vary  across  countries.  In  section  4.3  we  summarise  our  conclusions  about  the 
differences  in  national  transition  systems  which  explain  this  variation.  In  section  4.4 
we  explore  more  specific  issues:  the  role  of  education  in  the  transition  process,  the 
relative  importance  of  general  and  vocational  education,  and  the  nature  of  inequalities 
among  young  people.  Finally,  in  section  4.5  we  summarise  our  conclusions  and 
recommendations  on  data  issues. 


O 

ERIC 


89 


96 


4.2  CHANGING  TRANSITION  PROCESSES  AND  OUTCOMES 


Throughout  the  European  Union,  participation  and  attainment  in  education  have  been 
expanding.  By  the  late  1990s,  the  proportion  of  25-29  year  olds  in  the  EU  with  no 
qualifications  beyond  lower-secondary  education  stood  at  30  per  cent,  and  the 
proportion  with  tertiary  qualifications  stood  at  more  than  20  per  cent.  These 
proportions  varied  across  countries:  for  example,  the  UK  and  southern  European 
countries  have  more  low-qualified  young  people  than  elsewhere.  However,  at 
secondary  level  country  differences  have  narrowed  over  time,  as  countries  like  Ireland 
and  Spain  with  formerly  low  levels  of  upper-secondary  attainment  caught  up  with  the 
others.  At  tertiary  level,  however,  there  is  less  evidence  of  convergence.  In  some 
countries  (notably  Austria  and  Italy),  tertiary  attainment  has  tended  to  lag  behind  the 
rest.  In  general,  the  tertiary  sector  has  tended  to  expand  faster  in  countries  with  a less 
developed  vocational  education  at  upper-secondary  level. 

As  a result  of  these  trends,  young  people  now  enter  the  full-time  labour  market  with 
higher  qualification  levels  and  at  a later  age  than  previously.  However,  there  are  still 
wide  differences  within  and  between  countries  in  the  age  at  which  people  enter  the 
labour  market,  even  among  young  people  with  the  same  educational  level. 

In  some  European  countries  only  a minority  of  young  people  experience  relatively 
stable  employment  during  their  first  few  years  in  the  labour  market.  For  many  young 
people  these  years  are  characterised  by  unemployment  or  unstable  employment,  with 
a series  of  moves  between  different  statuses  such  as  education,  unemployment, 
inactivity,  youth  programmes,  military  service  and  part-time,  temporary  or  permanent 
jobs,  or  combinations  of  these.  However,  typical  transition  patterns  vary  across 
countries. 

In  most  countries  unemployment  rates  are  much  higher  among  young  people  entering 
the  labour  market  than  among  older  workers.  Unemployment  risks  decline  with  the 
length  of  time  in  the  labour  market.  However,  young  people  who  remain  unemployed 
for  a long  period  face  reduced  chances  of  obtaining  a job.  These  patterns  vary  across 
countries  as  does  the  overall  youth  unemployment  rate.  However,  comparisons  are 


affected  by  the  way  in  which  youth  unemployment  is  measured.  In  countries  with 
high  participation  in  education,  youth  unemployment  may  be  high  when  expressed  as 
a percentage  of  young  people  in  the  labour  force,  but  relatively  low  as  a proportion  of 
the  total  age  group.  On  the  former  definition,  the  EU  unemployment  rate  for  1995-97 
was  around  25  per  cent  for  16-21  year  olds,  declining  with  age  to  below  10  per  cent  at 
age  33.  As  a percentage  of  the  age  group,  unemployment  rose  to  around  14  per  cent  at 
age  21  and  then  declined  more  slowly,  falling  below  10  per  cent  at  age  28. 

A growing  proportion  of  young  people  pass  through  intermediate  or  mixed  statuses 
that  are  neither  wholly  in  education  nor  wholly  in  the  labour  market.  These  include 
training  or  employment  programmes  for  young  people.  Most  of  these  programmes 
were  introduced  as  a response  to  high  youth  unemployment.  In  some  countries  they 
retain  this  function  but  in  others  they  have  been,  to  varying  degrees,  assimilated  into 
the  initial  training  system.  They  also  vary  widely  in  scale  and  in  content,  some 
offering  work  experience  only  and  others  offering  structured  training. 

Other  intermediate  statuses  include  ‘dual’  statuses  which  combine  education  and 
work.  Across  the  EU  in  the  mid  1990s,  2 per  cent  of  the  whole  employed  labour  force 
(of  all  ages)  were  students  in  initial  education  who  also  had  jobs;  a further  2 per  cent 
were  apprentices;  and  6 per  cent  were  workers  attending  continuing  education.  The 
first  two  of  these  categories  were  heavily  concentrated  among  young  people;  the  third 
category  (studying  workers)  was  more  evenly  distributed  across  age  groups  with  the 
highest  representation  among  people  in  their  20s.  Apprenticeship  was  most  common 
in  Germany,  the  UK,  Denmark  and  Austria.  The  number  of  working  students,  and  the 
number  of  studying  workers,  tended  to  be  much  higher  in  northern  countries 
(Denmark,  Finland,  the  Netherlands,  the  UK  and  Sweden)  than  elsewhere.  All  three 
dual  status  categories  tend  to  be  less  common  in  southern  European  countries. 

A growing  proportion  of  young  people  spend  time  in  ‘non-standard’  forms  of 
employment,  such  as  part-time  or  temporary  jobs.  Once  again,  the  extent  of  these 
varies  very  widely  across  countries.  In  some  countries,  a large  proportion  of  young 
people  take  one  or  more  temporary  jobs  before  they  find  permanent  employment;  in 
other  countries,  temporary  jobs  are  a relatively  insignificant  feature  of  the  youth 
labour  market.  There  is  nevertheless  a general  trend  for  young  people  to  move  from 


less  secure  to  more  secure  jobs  as  they  grow  older.  They  also  tend  to  move  from  jobs 
with  lower  to  higher  occupational  status,  although  the  extent  and  nature  of 
occupational  mobility  also  varies  across  countries. 

It  is  arbitrary  to  identify  any  one  status  as  the  ‘final’  outcome  of  transition.  But  when 
- and  if  - young  people  find  full-time  and  relatively  stable  employment,  they  are 
more  likely  than  adults  to  work  in  clerical  or  sales  occupations  or  in  certain  service- 
sector  industries  such  as  hotels  and  restaurants. 

Finally,  these  processes  and  outcomes  of  transition  vary  for  males  and  females,  for 
young  people  from  different  social  backgrounds,  and  for  young  people  with  different 
levels  and  types  of  attainment  in  initial  education.  We  discuss  these  differences 
further  in  section  4.4  below.  While  variations  by  gender,  background  and  level  are 
found  everywhere,  their  precise  form  may  vary  between  countries. 

These  changing  characteristics  of  the  transition  from  education  to  work  have 
implications  for  guidance  and  information  services,  which  become  more  important 
with  the  longer  duration  and  greater  complexity  of  the  transition  process.  They  also 
have  clear  implications  for  policy-making  structures.  If  the  various  ‘statuses’  involved 
in  the  transition  are  the  responsibility  of  different  ministries  or  agencies  at  national 
level  (such  as  Education,  Training,  Employment,  Social  Security),  co-ordination 
among  them  is  essential. 

4.3  CLASSIFYING  TRANSITION  SYSTEMS 
4.3.1  The  continuum  of  transition  systems 

In  our  brief  summary  of  transition  processes  and  outcomes  we  have  drawn  attention  to 
the  fact  that  they  vary  across  countries.  What  are  the  main  types  or  dimensions  of 
national  transition  systems  which  explain  this  variation? 

In  developing  our  initial  conceptual  framework  (see  Chapter  3),  we  identified  several 
dimensions  of  variation  in  national  transition  systems,  and  suggested  that  many  of 
these  were  correlated  and  could  be  represented  as  a single  continuum.  At  one  end  of 


this  continuum  are  countries  such  as  Germany  with  strong  occupational  labour 
markets  (OLMs),  standardised  and  track-differentiated  education  systems,  and  strong 
links  between  education  and  the  labour  market  (often  through  apprenticeship 
systems).  At  the  other  end  of  the  continuum  are  countries  dominated  by  internal 
labour  markets  (ILMs),  with  less  standardised  and  less  differentiated  education 
systems,  weaker  links  between  education  and  the  labour  market  and  little  formal 
work-based  training.  Examples  of  the  latter  type  of  country  include  the  United  States 
and,  within  Europe,  Ireland  and  Scotland,  except  that  (like  other  EU  countries)  their 
education  systems  are  more  standardised  and  linkages  between  education  and  the 
labour  market  are  characterised  by  relatively  strong  market  signals. 

For  practical  reasons,  many  of  the  project  analyses  either  focused  on  a single 
dimension  of  variation  in  transition  systems  rather  than  on  the  continuum  as  a whole, 
or  they  used  categorical  rather  than  continuous  measures  to  classify  systems. 
Nevertheless,  the  analyses  confirm  the  value  of  the  conceptual  approach  adopted  by 
the  CATEWE  team.  The  dimensions  of  variation  in  transition  systems  are  indeed 
found  to  be  correlated,  and  countries’  transition  systems  tend  to  be  distributed  along  a 
continuum  such  as  that  described  above.  Moreover,  this  continuum  helps  to  explain 
national  variation  in  transition  processes  and  outcomes.  For  example,  in  countries  at 
the  former  end  of  the  continuum,  characterised  by  OLMs  and  strong  linkages  between 
education  and  the  labour  market: 

• youth  unemployment  rates  tend  to  be  lower,  and  the  ratio  of  youth  to  adult 
unemployment  rates  is  lower; 

• more  young  people  enter  the  labour  market  with  vocational  upper-secondary 
education,  and  somewhat  fewer  enter  with  tertiary  level  qualifications; 

• the  process  of  entry  to  the  labour  market  tends  to  be  smoother;  and, 

• initial  employment  is  less  likely  to  be  in  secondary-sector  jobs,  and  employment 
tends  to  be  more  stable. 

Most  of  these  differences  are  broadly  consistent  with  the  theoretical  ideas  underlying 
our  hypothesised  continuum.  However,  the  very  fact  that  the  component  dimensions 
of  this  continuum  are  correlated  makes  it  difficult  to  distinguish  empirically  among 
them,  and  among  the  different  theories  associated  with  them,  on  the  basis  of 


93 


•f  r\  r\ 


comparisons  of  a small  number  of  European  countries.  The  ‘intensive’  approach  to 
comparisons  (see  chapter  3)  deals  with  the  problem  raised  by  a small  sample  of 
countries  by  testing  multiple  predictions  based  on  its  theoretical  starting  point. 
However  existing  theories  tend  to  focus  on  one  dimension  at  a time.  We  need 
theoretical  perspectives  which  take  account  of  the  interrrelationships  of  different 
dimensions,  and  in  particular  a theory  which  links  the  educational  and  labour-market 
dimensions  of  variation  in  transition  systems. 

However,  while  our  research  confirms  the  utility  of  representing  transition  systems  in 
terms  of  the  continuum  described  above,  it  also  reveals  the  limits  of  the  continuum  as 
a single  explanatory  tool.  At  least  four  further  issues  need  to  be  taken  into  account. 

4.3.2  Other  dimensions  of  variation  in  transition  systems 

In  the  first  place,  the  research  shows  that  other  dimensions  of  transition  systems, 
which  vary  independently  of  the  overarching  continuum,  are  important.  These  include 

• the  strength  of  labour-market  regulation,  which  affects  the  extent  to  which  new 
entrants  are  'outsiders'  in  the  labour  market  and  consequently  influences  their 
employment  and  unemployment  chances.  Sometimes  the  influence  is  indirect,  if 
strong  regulation  of  the  adult  market  is  accompanied  by  a less  regulated  sector  for 
young  people; 

• the  extent  to  which  educational  achievements  are  differentiated  ‘vertically’  by 
formal  grades  or  levels.  This  may  be  especially  important  in  systems  where  track- 
differentiation  is  weak; 

• family  structure,  and  associated  with  this  the  state  institutions  (such  as  welfare 
regimes)  which  incorporate  assumptions  about  family  obligations,  and  other  social 
institutions,  such  as  housing  markets,  which  may  underpin  these  assumptions. 
Especially  in  Mediterranean  or  southern  European  countries,  family  support 
appears  to  influence  the  kinds  of  jobs  which  young  people  looking  for  jobs  are 
prepared  to  accept  and  the  length  of  time  they  are  prepared  to  wait  before 
accepting  a job; 

• more  generally,  aspects  of  the  supply  side  of  the  youth  labour  market  - young 
people’s  preferences,  transition  norms,  expectations  about  dependence  and 


94 


independence  and  the  age  of  transitions  to  adulthood.  Our  initial  conceptual 
framework  tended  to  assume  that  these  aspects  were  constant  across  all  countries, 
so  that  differences  could  be  attributed  to  features  of  the  demand  side  and  of 
educational  and  labour-market  institutions. 

4.3.3  Country-specific  features 

Second,  not  only  are  there  additional  dimensions  but  national  transition  systems  may 
have  features  which  are  specific  to  the  country  concerned  and  not  easily  reducible  to 
the  general  dimensions  of  the  conceptual  framework. 

For  example,  Schroder  (2000)  finds  variations  in  the  scale,  target  group  and  function 
of  youth  programmes  across  the  five  countries  in  the  school  leavers'  survey  dataset. 
While  these  variations  are  partly  explained  in  terms  of  general  dimensions  of 
transition  systems  (labour-market  regulation  and  the  strength  of  linkages  between 
education  and  the  labour  market),  they  sometimes  require  more  specific  explanations. 
For  example,  Scottish  youth  training  programmes  do  not  conform  to  Schroder’s 
theoretical  model,  because  they  have  been  developed  into  a mass  training  scheme  for 
employed  as  well  as  unemployed  young  people,  a fact  which  in  turn  calls  for  a more 
contextualised,  country-specific  explanation. 

In  chapter  3 we  described  three  methodological  approaches  to  comparison.  The 
project  has  used  the  third  of  these,  the  ‘interpretive’  approach,  to  identify  such 
country-specific  features.  An  example  is  the  study  by  Hartkamp  and  Rutjes  (2000)  of 
apprenticeship  in  four  countries.  In  the  Netherlands  and  France  apprenticeships  are 
more  parallel  to  upper-secondary  education,  are  less  selective  and  cover  a broader 
range  of  occupations  than  in  Scotland  and  (especially)  Ireland.  In  other  cases, 
comparisons  may  point  to  different  institutional  arrangements  which  perform  the 
analogous  function  - notably,  the  different  educational  mechanisms  which  produce  a 
vertical  differentiation  of  school  leavers,  respectively  track  placement,  certification, 
grades,  institution  type  or  school. 


4.3.4  Internal  differentiation  within  transition  systems 


A third  limitation  of  the  ‘single  continuum’  model,  and  more  generally  of  the  use  of 
dimensions  or  typologies  to  classify  education  systems,  is  that  different  parts  of  the 
same  system  may  belong  at  different  points  on  the  same  dimension  or  in  different 
categories  of  the  typology.  For  example,  the  extent  of  ‘standardisation’  may  vary 
across  different  sectors  of  a system.  Allmendinger’s  (1989)  paper  introducing  this 
concept  distinguished  among  the  standardisation  of  primary  and  secondary  schooling, 
of  higher  education  and  of  vocational  training  respectively  (although  her  hypotheses 
tended  to  treat  standardisation  as  a system-wide  concept).  Similarly,  the  strength  of 
labour-market  linkages,  or  the  extent  of  track  differentiation,  may  vary  across  the 
sectors  or  stages  of  an  education  system. 

Many  researchers  have  dealt  with  this  problem  by  focusing  on  a ‘critical  sector’  of 
education  or  training  which  is  assumed  to  define  the  system  as  a whole.  For  example, 
Germany  or  Denmark  are  treated  as  ‘dual  system’  countries  and  France  or  Sweden  as 
countries  with  school-based  vocational  education.  Often  this  assumption  is  implicit. 
Iannelli  and  Raffe  (2000)  propose  that  vocational  upper-secondary  education  is  the 
‘critical  sector’  in  this  sense  and  explore  the  implications  of  defining  a system  in 
terms  of  the  relationships  of  this  sector  with  other  parts  of  the  education  system  and 
with  the  labour  market.  They  hypothesise  that  these  relationships  may  follow  either  of 
two  ‘logics’  - one  in  which  vocational  upper-secondary  education  is  oriented  to 
higher  education  and  to  other  parts  of  the  education  system,  and  one  in  which  it  is 
more  closely  linked  to  the  labour  market.  They  find  some  evidence  to  support  their 
hypotheses,  but  given  the  data  limitations  the  evidence  is  not  conclusive. 

4.3.5  Changes  in  transition  systems 

The  fourth  limitation  of  the  notion  of  a single  explanatory  continuum  of  transition 
systems  is  that  the  characteristics  of  transition  systems  themselves  are  not  fixed.  For 
example,  even  a relatively  focused  policy  change,  such  as  the  withdrawal  of  benefit 
entitlements  from  unemployed  16  and  17  year  olds  in  Scotland  in  1988,  may  have  a 
significant  effect  on  the  process,  and  perhaps  also  the  outcomes,  of  transition.  In 
Ireland,  new  vocational  programmes  such  as  Post-Leaving  Certificate  (PLC)  courses 


have  had  a positive  impact  on  the  employment  chances  of  school  leavers,  in  apparent 
contradiction  to  the  representation  of  Ireland  as  an  ILM  country  with  weak  linkages 
between  education  and  the  labour  market,  where  we  would  expect  vocational 
education  to  have  relatively  poor  labour-market  returns.  Although  our  research  does 
not  reveal  the  reasons  for  these  programmes’  success,  it  is  likely  to  be  related  to  the 
provision  of  occupationally-specific  skills  within  PLC  courses  and,  perhaps,  to  recent 
radical  changes  in  the  Irish  economy  and  labour  market. 

As  the  transition  process  itself  changes,  new  dimensions  of  transition  systems  become 
important.  For  example,  rising  levels  of  educational  participation  and  the  prolongation 
of  the  transition  period  make  the  family  and  household  structure  and  the  support  that 
they  provide  to  young  people  in  education  and  seeking  employment  more  critical. 
Once  again,  closer  examination  of  Southern  European  countries  may  provide 
theoretical  lessons  that  are  of  wider  relevance. 

4.3.5  Overview 

The  four  ‘limitations’  listed  above  constitute  a challenge,  not  only  to  the  empirical 
proposition  that  a single  continuum  of  transition  systems  accounts  for  much  of  the 
cross-national  variation  in  transition  processes  and  outcomes,  but  also  to  the 
conceptual  and  theoretical  position  underlying  that  proposition.  On  the  one  hand,  they 
challenge  the  notion  of ‘dimensions’  of  transition  systems,  either  because  each  system 
has  features  that  are  sui  generis  and  not  reducible  to  a finite  number  of  general 
dimensions,  or  because  each  system  is  itself  heterogeneous  and  different  stages  or 
sectors  may  be  at  very  different  points  along  the  same  dimension.  On  the  other  hand, 
they  challenge  the  notion  of  transition  system  as  the  fixed,  independent  variable  of  the 
analysis.  Not  only  may  transition  systems  change,  but  they  may  do  so  in  response  to 
changes  in  the  ‘dependent  variable’,  in  the  transition  process  itself. 

The  conceptualisation  of  transition  systems  as  dimensioned,  stable  and  determining 
thus  contrasts  with  a societal  approach  which  emphasises  the  interaction  between 
different  parts  of  the  process  and  the  distinctive  logic  which  governs  each  country’s 
arrangements.  Our  project  has  been  based  on  the  assumption  that  both  approaches  are 
necessary  for  a full  understanding  of  transitions. 


4.3.6  Policy  implications 


The  main  policy  implication  from  our  analysis  of  transition  systems  is  that  policies  for 
the  transition  from  education  to  work  must  be  designed  in  relation  to  the  particular 
transition  system  in  which  they  are  to  be  introduced.  Policies  which  are  effective  in 
one  type  of  transition  system  are  not  necessarily  effective  in  another.  The  trend  in 
cross-national  policy  thinking  is  moving  away  from  identifying  institutions  or  policies 
which  are  expected  to  be  universally  effective,  towards  identifying  the  key  features 
(or  conditions)  of  effective  transition  systems.  These  conditions  might  be  achieved 
through  different  policies  in  different  transition  systems.  Our  research  endorses  this 
trend,  which  is  reflected  (for  example)  in  the  final  report  of  the  OECD’s  (2000) 
Thematic  Review  on  the  Transition  from  Initial  Education  to  Working  Life. 

The  implication  for  policy-makers  at  EU  level,  therefore,  is  that  uniformity  of  policy 
across  the  diverse  transition  systems  of  the  EU  is  not  necessarily  desirable.  For 
national  policy-makers,  it  cannot  be  assumed  that  policies  can  be  'borrowed'  from 
other  national  systems  and  prove  equally  effective.  To  the  extent  that  policy 
borrowing  is  ever  appropriate,  it  is  most  likely  to  be  effective  among  countries  with 
similar  types  of  transition  systems. 

This  is  not  to  say  that  countries  should  never  attempt  to  reform  their  transition 
systems,  and  here  the  concept  of  dimensionality  may  be  a guide.  Our  notion  of  a 
‘single  continuum’  draws  attention  to  the  interrelated  nature  of  the  dimensions 
discussed  in  chapter  3,  even  if  it  also  signals  the  need  for  further  research  and 
theoretical  development.  Some  dimensions  are  more  easily  changed  through  policy 
intervention  than  others. 

4.4  SPECIFIC  ISSUES  IN  THE  ANALYSIS  OF  TRANSITIONS 
4.4.1  Educational  level  and  its  relationship  with  exclusion 

The  level  of  education  attained  is  a crucial  variable  shaping  individual  education-to 
work  transitions.  In  numerous  analyses,  based  on  both  SLS  and  LFS  databases,  the 


project  has  established  unequivocal  evidence  for  the  strong  relationship  between  level 
of  education  and  core  transition  outcomes.  The  higher  the  individual  level  of 
education,  the  lower  the  risk  of  unemployment,  the  faster  initial  jobs  will  be  found, 
which  will,  in  addition,  be  more  stable  and  in  more  prestigious  occupations  and 
industries,  pay  a better  salary,  and  offer  better  opportunities  for  further  training, 
among  other  things.  Individual  investment  in  education  and  training  pays  off  and 
there  is  little  evidence  of  fundamental  differences  between  European  countries, 
despite  some  cross-national  variation  which  has  been  discussed  in  chapter  3 above.  In 
other  words,  those  leaving  the  educational  system  at  a very  early  stage,  that  is,  from 
lower  secondary  education  or  even  without  properly  completing  compulsory 
education,  are  very  likely  to  face  considerable  difficulties  in  the  early  stages  of  their 
careers.  And  although  the  project  itself  has  focused  on  the  effects  of  education  in  the 
very  first  years  on  the  market,  it  is  well-known  from  other  research  that  such  initial 
difficulties  will  often  continue  into  later  career  stages  as  well  (see,  for  example, 
Hammer,  1 997).  The  disadvantages  faced  by  those  entering  the  market  with  relatively 
little  education  and  training  seems  a robust  and  unequivocal  finding  across  all 
European  Union  countries  over  the  historical  period  covered  by  the  project’s  data 
sources. 

Consequently,  concerns  for  appropriate  policy  measures  have  been  widespread  for  a 
long  time,  both  at  the  European  level  and  within  EU  member  states.  In  fact,  there  are 
a number  of  additional  project  findings  which  stress  the  need  to  maintain  strong 
policy  orientations  toward  the  lowest  qualified.  First,  we  have  produced  considerable 
evidence  for  a strong  macroeconomic  component  in  transition  outcomes.  Labour 
market  outcomes  among  those  entering  the  labour  market  are  heavily  affected  by 
aggregate  economic  conditions,  much  more  so  than  is  the  case  for  more  experienced 
workers.  Least  qualified  market  entrants  are  found  to  be  the  most  vulnerable  to 
cyclical  swings  in  economic  activity  (see  also  Blanchflower  and  Freeman,  2000). 
Steady  economic  growth  across  European  countries  is  likely  to  mean  that  transition 
issues,  particularly  youth  unemployment,  will  be  given  less  policy  attention. 
However,  likely  future  changes  in  macroeconomic  conditions  will  have  particular 
implications  for  young  people  entering  the  labour  market  and  will  thus  raise  issues  for 
policy-makers.  For  this  reason,  it  is  crucial  that  we  now  take  stock  of  alternative  ways 


99 


to  successfully  integrate  young  people  into  the  labour  market  and  begin  to  consider 
potential  institutional  reform  to  better  achieve  such  integration. 

But  are  qualifications  and  skills  the  answer  to  this?  And  if  so,  which  skills?  More 
advanced  qualifications  or  completely  different  skills,  better  training  or  more 
specialised  preparation  for  working  life,  reform  of  initial  education  and  training  or 
expansion  of  adult  and  further  education  systems?  Of  course,  our  project  cannot  claim 
to  be  able  to  provide  definite  answers  to  any  of  these  questions.  Still,  some  modest 
conclusions  are  defensible.  A first  and  tentative  result  is  an  indication  that  those  at  the 
bottom  end  of  educational  qualifications  are  unable  to  benefit  from  expanding 
employment  opportunities  in  modem(ising)  sectors  of  the  economy.  If  this  is 
supported  by  future  research,  the  implications  are  highly  significant.  In  fact,  this 
might  be  a first  signal  that  - despite  recent  and  current  reforms  updating  the 
curriculum  of  compulsory  education  in  many  countries  - the  actual  skill  contents  of 
these  tracks  may  increasingly  fall  short  of  the  needs  of  structurally  changing  labour 
markets  and  a stable  working  life  therein  (see  also  McIntosh  and  Steedman,  1999). 
Such  trends  should  not  be  exaggerated  prematurely,  but  if  this  indication  substantiates 
itself  further,  then  policy  action  to  prevent  early  school  drop-out  and  foster  even 
increasing  - initial  and/or  second  chance  - participation  in  advanced  forms  of 
education  and  training  might  well  be  warranted.  This  is  not  to  imply  that  EU  member 
states  have  proven  inattentive  to  the  issue  in  the  recent  past  but  to  emphasise  that 
providing  all  young  people  with  an  adequate  skill  basis  (including  literacy  and 
numeracy)  for  their  working  lives  is  and  remains  a serious  issue,  which  deserves 
special  attention  in  the  context  of  rapidly  changing  economies. 

Having  said  this,  it  should  probably  be  added  that  our  results  leave  us  quite  confident 
that  further  educational  expansion  is  not  very  likely  to  prove  as  detrimental  to  young 
people’s  labour  market  outcomes  and  returns  to  educational  investments  as  is  often 
assumed  in  current  research.  True,  education  has  the  quality  of  a positional  good,  so 
that  its  value  on  the  market  is  partly  determined  by  just  how  exclusive  any  particular 
qualification  is.  In  addition,  concerns  about  the  incidence  of  over-education  have  been 
most  prominent  among  education  and  labour  market  scholars.  But  according  to  our 
empirical  results,  the  net  decline,  if  any,  in  returns  to  education  has  been  quite  small 
for  most  educational  groups,  despite  remarkable  expansion  of  education,  notably  at 


the  tertiary  level.  This  is  not  to  say  that  educational  expansion  will  never  result  in  a 
devaluing  of  certain  credentials.  But  apparently,  for  most  of  the  1980s  and  1990s, 
such  downgrading  tendencies  have  been  offset  by  parallel  changes  in  the  structure  of 
labour  demand,  and  young  people  have  been  able  to  benefit  from  these  changes  to  a 
considerable  degree.  Hence,  these  results  seem  more  consistent  with  an  interpretation 
that  stresses  that  considerable  and  productive  additional  skills  are  conveyed  by  more 
advanced  education,  which  allows  individuals  to  benefit  from  employment 
opportunities  in  the  expanding  sectors  of  the  labour  market,  or  even  contribute  to  the 
speed  of  structural  changes  in  the  economy  itself.  In  order  to  generalise  these  findings 
further  and  to  understand  the  nature  of  these  empirical  linkages,  further  study  of  the 
inter-relatedness  of  educational  expansion  and  labour  market  changes  seems  to  be 
warranted. 

In  general,  there  can  thus  be  little  doubt  from  our  results  that  increased  levels  of 
education  give  individuals  a better  start  into  a more  promising  working  life.  If  this  is 
correct,  then  the  main  policy  question  becomes:  why  do  some  individuals  leave  the 
educational  system  with  low  and  potentially  insufficient  levels  of  education,  and  how 
might  this  be  remedied  in  the  future?  It  cannot  be  claimed  that  we  have  treated  this 
question  sufficiently  within  the  lifetime  of  the  project,  but  there  are  some  results 
which  appear  relevant  to  it,  which  should  help  to  formulate  adequate  questions  in 
future  research.  As  these  are  closely  linked  to  the  role  played  by  vocational  training, 
they  will  be  discussed  below. 

4.4.2  Vocational  education  and  entry  into  working  life 

Large-scale  systems  of  vocational  education  have  long  been  considered  and  utilised  as 
the  main  means  of  providing  relevant  skills  to  non-academically  oriented  young 
people  in  many  countries.  In  fact,  as  far  as  can  be  told  from  the  data,  substantial 
proportions  of  those  young  people  not  entering  academic-bound  educational  tracks 
participate  in  some  form  of  vocational  training  before  entering  the  labour  market  or 
over  the  initial  years  after  leaving  the  school  system.  Given  huge  variation  in 
institutional  arrangements  for  providing  vocational  training,  it  is  not  surprising  that 
the  distribution  of  young  people  across  various  types  of  training  varies  considerably 
across  European  countries.  Institutionally  diverse  as  they  are,  national  systems  of 


vocational  training  differ  according  to  the  nature  of  vocational  specialisation  offered 
and  the  number  of  occupations  trained  for,  the  level  of  entry  qualifications  required, 
the  dominance  of  school-based  versus  dual  forms  of  training  which  combine  formal 
training  and  work,  the  extent  and  nature  of  provision  for  work  experience  during 
training,  or  in  the  extent  and  nature  of  direct  or  indirect  employer  involvement  in  both 
training  design  and  provision,  among  others.  In  addition,  the  dividing  line  between 
vocational  training  and  active  labour  market  policies  targeted  at  young  people  is  often 
difficult  to  draw.  Some  of  the  more  ambitious  and  long-running  training  schemes  (like 
those  in  the  United  Kingdom)  are  significant  components  of  the  training  system  and 
have  long  ago  abandoned  the  primary  objective  of  providing  work  experience  to 
unemployed  youth. 


This  cross-national  variation  in  institutional  settings  provides  ample  opportunity  to 
assess  the  effects  of  different  types  of  vocational  education  under  different  contextual 
conditions  and  with  respect  to  different  aspects  of  transition  outcomes.  The  main 
questions  to  be  settled  are:  What  are  the  effects  of  completing  vocational  training  in 
comparison  to  leaving  from  more  general  educational  tracks?  Does  vocational  training 
represent  a reasonable  educational  investment,  at  least  in  terms  of  initial  employment 
returns  to  it?  Or  does  vocational  training  imply  a risk  of  diversion  by  trapping  young 
people  in  less  attractive  and  menial  employment  careers,  and  with  less  access  to 
higher  levels  of  education  subsequently?  And  are  the  effects  of  vocational  training 
(near)  universal  or  are  there  peculiar  institutional  features  of  labour  markets  which 
affect  the  role  of  vocational  training?  And  coming  back  to  the  preceding  section,  if 
vocational  training  does  provide  labour  market  advantages,  does  it  seem  likely  that  it 
represents  a sufficiently  attractive  route  to  people  who  would  otherwise  leave  the 
school  system  at  relatively  early  stages? 


ERIC 


The  answer  to  these  questions  is  first  and  foremost  a question  of  the  standard  of 
comparison.  From  what  we  can  tell  from  the  project's  results,  vocational  training  does 
facilitate  labour  market  entry  and  provides  access  to  better  employment  opportunities 
than  are  available  to  those  having  left  education/training  systems  with  lower 
qualifications.  Compared  to  young  people  entering  the  labour  market  with  more 
academically-oriented  upper  secondary  qualifications,  vocationally  qualified  leavers, 
in  general,  do  not  achieve  substantially  different  unemployment  or  employment 


102 


10.9 


outcomes.  If  anything,  apprentices  tend  to  have  somewhat  lower  unemployment  rates 
but  also  somewhat  less  favourable  employment  outcomes.  In  contrast,  labour  market 
outcomes  for  those  leaving  school-based  vocational  training  and  those  from  general 
tracks  at  upper  secondary  level  are  typically  quite  similar. 

These  results  showing  close  similarities  in  labour  market  effects  of  vocational  training 
across  the  set  of  European  economies  are  quite  remarkable  in  themselves.  Leavers 
from  apprenticeships  or  similar  forms  of  combined  work-training  contracts  have 
considerable  advantages  in  securing  relatively  stable  employment  positions  quickly. 
This  relationship  apparently  holds  for  programmes  as  institutionally  diverse  as  those 
run  in  countries  such  as  Austria,  Denmark,  Germany,  France,  the  Netherlands,  or  the 
United  Kingdom.  Such  similarity  strongly  suggests  that  no  particular  macro- 
institutional  context  is  necessary  for  apprenticeships  to  work1;  apprenticeships  are 
mainly  successful  because  they  provide  a sheltered  work  contract  with  a particular 
employer.  This  design  feature  is  sufficient  to  generate  smoother  transition  patterns  in 
different  market  contexts  because  it  opens  up  the  option  of  continued  employment  at 
the  training  company  after  completing  the  training  period.  It  is  this  feature  which  is 
mainly  responsible  for  low  unemployment  rates  among  apprentices  in  the  early  stages 
of  labour  market  careers,  compared  with  leavers  from  both  general  and  vocational 
tracks  within  upper  secondary  education. 

Context-independent  effects  of  apprenticeships  are  also  evident  in  relation  to 
occupational  outcomes:  we  have  gained  no  evidence  that  apprentices  in  the  more 
occupationalised  contexts  of  Austria,  Denmark,  Germany,  and  the  Netherlands  have 
systematically  more  favourable  employment  outcomes  than  apprentices  in  other 
European  economies.  Comparing  European  countries  operating  apprenticeship 
systems,  the  empirical  picture  is  one  of  considerable  heterogeneity  among  countries 
with  strong  OLMs,  rather  than  any  clear-cut  contrast  between  these  countries  and  (for 


This  does  not  conflict  with  findings  that  installing  and  maintaining  apprenticeship  systems  is  greatly 
influenced  by  the  current  macro-institutional  context.  It  might  well  be  helpful  to  have  strong  trade 
unions,  a historical  legacy  of  vocational  training,  co-ordination  among  employers  and  strong  employer 
involvement  in  training  policy  for  developing  feasible  apprenticeship  tracks,  although  none  of  these 


example)  France,  the  United  Kingdom  or  Ireland.  In  fact,  the  average  apprentice  in 
Germany  is  found  to  have  the  most  positive  occupational  outcomes  across  European 
countries,  while  apprenticeships  in  the  Netherlands  and  Austria  have  the  least  positive 
outcomes.  What  is  at  stake  here  is,  of  course,  not  the  quality  of  the  programme  per  se, 
but  the  range  of  occupations  being  trained  for,  and  notably  the  reach  of  the  system 
into  the  service  sector,  both  of  which  are  considerably  larger  in  Germany  than  in  the 
two  other  countries  mentioned.  Clearly,  the  nature  of  occupations  trained  for  becomes 
closely  reflected  in  subsequent  initial  employment  experiences. 

While  the  effects  of  vocational  training  discussed  so  far  have  emerged  as  broadly 
common  to  the  set  of  European  countries,  there  are  two  cases  where  we  have  found 
training  effects  to  differ  systematically  between  countries,  and  both  relate  to  the 
effects  of  school-based  vocational  training  rather  than  apprenticeship  programmes. 
With  respect  to  both  pay  levels  and  unemployment  risks,  we  have  found  more 
favourable  outcomes  for  leavers  from  school-based  vocational  training  in  those 
countries  where  such  training  occurs  in  the  context  of  strong  OLMs  - notably  in  the 
Netherlands,  but  to  a lesser  extent  also  in  Austria  and  Germany.  It  appears  that  these 
qualifications  are  more  explicitly  recognised  by  employers  in  more  occupationalised 
systems.  For  policy  purposes,  this  is  likely  to  hold  the  implication  that  the  full  value 
of  any  expansion,  introduction  or  change  of  school-based  vocational  training  in  the 
context  of  weakly  occupationalised  systems  will  only  be  realised  after  some  time, 
presumably  because  employers  need  relatively  more  time  to  become  aware  of  the 
contents  of  such  training,  as  well  as  the  reliability,  adequacy  and  relevance  of  this 
educational  output  as  compared  to  systems  with  already  long-established  practices  of 
occupationally  specific  training. 

Much  of  the  foregoing  discussion  suggests  that  no  single  institutional  panacea  exists 
which  uniquely  generates  successful  and  smooth  pathways  into  the  labour  market. 
Indeed,  with  respect  to  many  labour  market  outcomes  studied,  those  for  leavers  from 
apprenticeships,  school-based  vocational  training  or  upper  secondary  general 
education  are  often  strikingly  similar  across  European  countries.  This  similarity  leads 


institutional  preconditions  for  the  existence  of  such  systems  strongly  affects  the  subsequent  nature  of 
its  stabilising  effects  on  labour  market  careers. 


104 


ill 


us  to  emphasise  the  importance  of  training  for  successful  labour  market  entry  per  se 
rather  than  the  institutional  nature  of  training  provision:  the  main  determinant  of 
initial  career  experiences  is  clearly  the  level  of  initial  education  and  training,  rather 
than  the  type  of  education.  This  also  suggests  that  providing  training  and  skills  to 
young  people  is  the  key  to  avoiding  transition  problems.  Thus,  providing  skills 
through  general  or  vocational  programmes  might  in  fact  be  seen  as  alternative  and 
potentially  complementary,  rather  than  as  antagonistic  and  mutually  exclusive, 
strategies.  The  benefits  of  adopting  or  further  expanding  either  route  of  training 
should  very  much  depend  on  its  contribution  to  the  final  levels  of  training  achieved  by 
young  people.  Before  considering  the  relative  merits  of  general  versus  vocational 
education  in  that  respect,  one  should,  however,  note  one  distinctive  feature  of 
apprenticeships:  even  though  labour  market  outcomes  among  apprentices  are  unlikely 
to  differ  dramatically  from  those  having  left  other  types  of  upper  secondary  education 
in  the  longer  perspective,  apprentices  are  likely  to  achieve  their  integration  into  the 
labour  force  with  considerably  lower  levels  of  initial  unemployment  as  well  as  fewer 
and  shorter  periods  of  job  search.  The  main  feature  of  apprenticeships  which  is  likely 
to  bring  about  such  favourable  initial  outcomes  is  the  established  relationship  with  a 
particular  employer  and  the  resulting  chances  of  continued  employment  after  training 
completion.  Training  tracks  lacking  this  design  feature  do  not  necessarily  imply 
different  labour  market  outcomes  by  the  age  of  25,  30  or  even  35,  but  it  is  likely  that 
individuals  will  experience  more  problems  in  securing  stable  employment  initially. 
Compared  to  school-based  training,  apprenticeships  do  have  a certain  safety  net 
effect. 


O 

ERIC 


But  if,  in  the  long  run,  different  modes  of  training  provision  achieve  relatively  similar 
labour  market  outcomes,  then  there  is  little  to  choose  between  institutional  forms  of 
training  provision  with  better  or  worse  outcomes,  but  rather  in  terms  of  their  inherent 
skill  formation  potential.  That  is,  initial  education  and  training  aimed  at  minimising 
transition  problems  would  be  well  advised  to  devise  a mix  of  general  education  and 
apprenticeship  or  other  types  of  vocational  training  which  maximises  individual  and 
cohort  skill  acquisition.  Effectively,  there  are  two  main  concerns  here:  training 
incentives  and  the  diversion  effects  of  vocational  training.  The  first  argument 
favouring  the  use  of  vocational  training  provision  would  emphasise  the  potentially 
lower  thresholds  to  skill  acquisition  through  vocational  training,  related  to,  for 


105 


112 


example,  the  more  applied  nature  of  the  skills  trained  and  the  provision  of  more 
concrete  pathways  into  the  labour  market.  Hence,  vocational  training  could  serve  as 
an  effective  means  to  achieve  high  progression  rates  beyond  compulsory  levels  of 
education  as  those  otherwise  entering  the  labour  market  without  any  additional 
training  might  find  a reasonably  attractive  training  alternative.  Supportive  to  this 
notion  are  project  findings  on  a positive  correlation  between  the  size  of  a country’s 
vocational  training  system  and  the  proportion  of  individuals  who  progress  beyond 
compulsory  levels  of  education,  as  well  as  higher  progression  rates  beyond 
compulsory  education  in  the  Netherlands  as  compared  to  other  SLS  countries. 

Contrary  to  this,  the  diversion  argument  stresses  the  fact  that  vocational  training 
might  inhibit  further  skill  formation.  Indeed,  even  if  vocational  and  general  training 
provide  formally  equal  access  to  further  training,  progression  rates  into  post- 
secondary and  tertiary  education  are  typically  much  lower  among  leavers  from 
vocational  training  compared  to  those  among  leavers  from  more  general  tracks.  To 
some  extent,  however,  this  outcome  will  reflect  the  different  orientations  of  young 
people  when  they  enter  these  tracks:  those  joining  vocational  tracks  are  more  likely  to 
plan  a relatively  early  entry  to  the  labour  market.  On  the  other  hand,  young  people’s 
orientations  about  their  subsequent  career  are  unlikely  to  be  completely  fixed  by  the 
age  of  16  or  18,  and  education  and  training  undergone  will  considerably  shape  and 
develop  these  orientations.  In  consequence,  the  more  practical  orientation  of 
vocational  training  can  be  assumed  to  contribute  to  narrowed  perceptions  of,  and 
expectations  about,  individual  career  options,  which  may  bring  about  a diversion  of 
academic  effort  and  potential. 

In  some  sense,  vocational  training  is  likely  to  maintain  a Janus-faced  appeal,  as  both 
incentive  and  diversion  effects  are  simultaneously  apparent  in  our  results  (see  also 
Shavit  and  Muller,  2000).  That  is,  on  the  one  hand,  vocational  training  might 
contribute  to  increased  training  participation  among  those  who  would  otherwise  not 
continue  into  upper  secondary  education.  On  the  other  hand,  however,  young  people 
in  vocational  training  are  less  likely  to  further  invest  in  education  and  training  than 
those  in  more  general  tracks.  Ultimately,  strongly  vocationally-oriented  systems  are 
thus  likely  to  achieve  particularly  high  levels  of  intermediate  skills.  These  do  have  the 
advantage  of  providing  more  adequate  skills  to  that  part  of  a cohort  who  would  not 


have  progressed  beyond  compulsory  education,  but  they  do  constrain  further 
educational  achievement  among  those  who  would  have  entered  more  academic  tracks 
at  the  upper  secondary  level  otherwise.  A satisfactory  empirical  assessment  of  both 
components  would  seem  a particularly  important  task  for  future  research.  In  the 
absence  of  such  research,  the  above  discussion  indicates  why  educational  policy 
priorities  differ  across  different  European  countries:  the  main  problem  in  systems  with 
strong  OLMs  is  likely  to  be  overcoming  the  diversion  effects  of  large-scale  systems  of 
vocational  training,  while  policy  makers  in  more  general  educational  systems  will  be 
more  concerned  about  achieving  higher  progression  rates  into  upper  secondary  or 
equivalent  training. 

4.4.3  Equity  issues  in  transitions 

Much  of  the  foregoing  discussion  of  school  to  work  transitions  has  focused  on 
comparing  'average'  transitions  across  different  countries  and/or  institutional  systems. 
However,  the  pathways  taken  by  different  groups  of  young  people  within  specific 
national  contexts  can  vary  markedly.  The  extent  to  which  this  diversity  is  structured 
by  gender,  ethnic  and  social  differences  has  been  the  subject  of  much  debate.  Some 
commentators  have  argued  that  individual  trajectories  from  school  to  the  labour 
market  have  become  more  diverse  but  at  the  same  time  less  bound  by  factors  such  as 
gender,  social  background  and  so  on.  Analyses  from  the  CATEWE  project  shed  some 
light  on  this  debate,  indicating  no  general  trend  towards  greater  individualisation  and 
revealing  the  persistence  of  gender,  social  background  and  national  origin  as 
important  influences  on  the  transition  process.  However,  these  differences  are  found 
to  be  constructed  in  distinct  ways  in  different  national  contexts  and  at  different 
periods  of  time. 

Gender 

Recent  decades  have  been  characterised  by  considerable  educational  expansion  across 
Europe,  particularly  among  young  women.  In  many  European  countries,  women  in 
the  older  cohorts  were  less  likely  than  men  to  attain  higher  level  qualifications. 
Among  younger  cohorts,  this  difference  has  diminished  and  in  many  countries  female 
attainment  of  upper  secondary  education  has  surpassed  that  of  men.  Similarly,  gender 
differences  in  the  attainment  of  tertiary  education  have  diminished  over  time,  with 


significantly  higher  attainment  levels  now  evident  among  females  than  males  in 
Portugal,  Belgium  and  Finland.  Within  secondary  education,  young  women  tend  to 
outperform  their  male  counterparts  in  examinations,  at  least  in  Ireland,  Scotland  and 
Sweden,  though  not  the  Netherlands. 

The  increasing  educational  participation  of  young  women  across  Europe  has  not  been 
accompanied  by  a significant  diminution  in  gender  segregation  across  different  types 
of  educational  routes.  Young  women  tend  to  be  less  likely  to  participate  in  school- 
based  vocational  education  than  young  men,  and  even  in  systems  with  high  levels  of 
such  provision  (such  as  the  Netherlands),  the  nature  of  participation  tends  to  be 
strongly  differentiated  by  gender.  Participation  in  apprenticeship  training  also  tends  to 
be  disproportionately  concentrated  among  males  across  European  countries.  Even  in 
countries  with  higher  female  participation  in  apprenticeships,  gender  differences 
persist  in  the  occupational  areas  to  which  such  apprenticeships  are  oriented.  In 
addition,  within  tertiary  education  women  are  over-represented  in  lower  level  and  less 
prestigious  programmes. 

What  are  the  implications  then  of  gender  differences  in  educational  attainment  for 
labour  market  transitions  among  young  people?  The  gender  distribution  of 
unemployment  varies  across  European  countries  with  no  evidence  of  an  'average' 
gender  difference  across  all  countries.  A commonality  across  countries  is  the 
existence  of  occupational  segregation  by  gender,  although  the  form  taken  by  this 
segregation  varies  cross-nationally.  There  is  tentative  evidence  that  segregation  is 
greater  in  systems  (such  as  the  Netherlands)  with  greater  track  differentiation  at 
school  level,  in  part  because  these  tracks  tend  to  channel  young  people  into  gendered 
occupational  fields.  However,  segregation  also  occurs  on  entry  to  the  labour  market 
since  even  young  women  and  men  with  the  same  level  and  type  of  education  tend  to 
end  up  in  very  different  occupational  and  industrial  niches.  On  average,  young  women 
tend  to  enter  higher  status  occupations  than  young  men.  However,  there  is  tentative 
evidence  that  they  receive  lower  returns  to  their  education  in  terms  of  status  and  pay 
than  men. 

Given  the  often  diverse  pathways  taken  by  young  women  and  men  in  their  transition 
from  education  to  the  labour  market,  it  is  clear  that  any  policy  interventions  cannot 


afford  to  be  'gender-blind'.  Furthermore,  since  the  nature  of  gender  differences  in 
education  and  labour  market  outcomes  varies  cross-nationally,  policy  also  needs  to  be 
sensitive  to  the  national  institutional  context.  It  appears  that  there  is  no  necessary 
relationship  between  the  type  of  education/training  system  and  the  extent  of  gender 
differences  in  educational  attainment.  There  is  tentative  evidence,  however,  that 
countries  with  higher  levels  of  provision  of  vocational  education  (either  within  school 
or  as  part  of  a 'dual'  arrangement)  tend  to  be  more  successful  in  promoting  educational 
attainment  among  young  males. 

What  appears  to  be  common  across  European  countries  is  that  educational  and 
occupational  segregation  by  gender  have  remained  remarkably  persistent  in  the  face 
of  the  introduction  of  equal  opportunity  legislation  and  positive  action  programmes.  It 
is  evident  that  information  is  needed  on  an  on-going  basis  on  both  the  institutional  and 
the  informal  factors  shaping  the  interaction  between  the  education  and  labour  market 
systems  as  well  as  on  formal  gender  differences  in  the  transition  process.  - 

Social  background 

Previous  research  has  indicated  the  persistence  of  social  class  inequalities  in 
educational  attainment  and  has  suggested  that  such  inequalities  are  likely  to  be  greater 
in  systems  with  earlier  selection  into  different  educational  tracks  (Shavit  and 
Blossfeld,  1993;  Muller  and  Shavit,  1998).  As  such,  social  class  background  is  likely 
to  have  at  least  an  indirect  effect  on  the  transition  process.  In  general,  there  is  very 
little  systematic  comparative  data  on  social  class  background  and  its  relationship  with 
educational  and  labour  market  outcomes  (OECD,  2000).  The  Labour  Force  Survey 
does  not  collect  such  information  and  some  national  transition  surveys  lack 
comparable  data  on  this  measure.  Analyses  for  a limited  range  of  countries  indicate 
the  persistence  of  the  relationship  between  family  background  and  educational 
attainment,  with  those  from  working-class  backgrounds  disproportionately  over- 
represented among  the  least  qualified.  Social  class  background  is  also  associated  with 
entry  to  further  education,  employment  chances  and  job  quality,  even  controlling  for 
prior  educational  qualifications.  It  is  recommended  that  information  on  socio- 
economic background  be  collected  on  an  on-going  basis  for  a wider  range  of  countries 
in  order  to  explore  the  relationship  between  social  class  background,  educational 
outcomes  and  labour  market  integration. 


National  origin 

Ethnicity  and  national  origin  have  been  neglected  areas  in  comparative  transitions 
research,  in  part  because  of  the  absence  of  systematic  data  on  the  topic.  Data  on  the 
national  origins  of  young  people  were  available  from  the  French  and  Swedish 
longitudinal  surveys.  Analyses  of  these  data  are  highly  revealing  of  the  significant 
role  of  national  origins  in  structuring  the  transition  process.  Those  bom  abroad  and 
those  with  two  parents  bom  aboard  are  at  a disadvantage  in  terms  of  educational 
attainment  compared  with  native-born  young  people.  This  disadvantage  is  more 
evident  in  France  which  has  a more  selective  educational  system.  Differences  are  also 
evident  in  early  labour  market  outcomes  with  higher  unemployment  risks  among 
immigrant  youth,  even  controlling  for  education.  Direct  labour  market  disadvantage  is 
more  evident  in  Sweden  where  education  plays  a less  clear-cut  signalling  role  than  in 
the  French  context.  It  is  interesting  to  note  that  the  way  in  which  differences  by 
national  origin  are  constructed  varies  across  countries;  in  France,  the  group 
experiencing  greatest  disadvantage  are  young  people  of  North  African  origin  while  in 
the  Swedish  case  this  comprises  those  of  Asian  (Turkish)  Latin  American  and  other 
African  origin. 

The  relevance  of  ethnic/national  differentiation  to  transitions  may  vary  across 
European  countries.  However,  it  is  clear  that  information  is  needed  on  national  origins 
in  order  to  explore  relative  disadvantage  in  different  educational  and  labour  market 
contexts.  Such  data  need  to  reflect  the  national  context,  exploring  the  ethnic/national 
distinctions  which  are  most  appropriate  in  that  context. 

In  general,  greater  attention  should  be  given  to  the  development  of  equity  measures  in 
education  and  transition  outcomes  which  are  at  the  same  time  comparable  across 
countries  and  sensitive  to  the  realities  of  the  national  context.  In  addition  to  social 
class  and  ethnicity,  other  potential  sources  of  differentiation  in  young  people's 
pathways  should  be  explored.  For  example,  there  has  been  almost  no  attention  given 
to  the  transition  process  experienced  by  young  people  with  disabilities. 


4.5  DATA  ISSUES  AND  RECOMMENDATIONS 


The  experience  of  the  CATEWE  project  has  highlighted  gaps  in  existing  transition 
research,  the  strengths  and  weaknesses  of  existing  data  on  the  topic  and  issues  to  be 
considered  in  further  work. 

4.5.1  Areas  for  further  research 

Four  main  gaps  are  evident  in  existing  research.  In  the  first  place,  comparatively  little 
is  known  about  employer  strategies  in  the  labour  market,  particularly  as  they  impact 
on  young  people.  At  present,  researchers  tend  to  make  inferences  about  employer 
preferences  and  practices  on  the  basis  of  the  actual  pattern  of  employment  and 
unemployment  rather  than  from  direct  evidence.  Little  is  known  about  the  factors 
which  underlie  employers'  recruitment  decisions  or  the  implications  of  changing 
employment  practices  for  the  hiring  (or  firing)  of  young  people.  In  the  face  of 
changing  educational  and  training  provision,  employers'  use  and  perceptions  of 
educational  'output'  has  been  given  little  attention.  This  area  of  research  is  likely  to 
prove  particularly  fruitful  in  assessing  the  likely  impact  of  educational  reforms  and 
changing  employment  practices. 

Secondly,  research  on  the  actual  transitions  made  by  young  people  needs  to  be 
complemented  by  studies  of  the  ways  in  which  young  people  experience  this  process. 
What  factors  shape  young  people's  decisions  about  the  routes  they  take  within  and 
beyond  the  educational  system?  What  are  their  goals  and  expectations?  Do  terms  such 
as  'over-education'  have  real  meaning  in  the  lives  of  young  people?  How  does  young 
people's  experience  of  paid  work  shape  their  family  and  interpersonal  relationships, 
and  vice  versa?  Although  there  have  been  several  research  studies  of  such  issues, 
relatively  few  have  been  cross-national  and  comparative  (e.g.  Evans  and  Heinz,  1996; 
Heinz,  1999).  Such  studies  are  crucial  to  an  adequate  understanding  of  (changes  in) 
transition  processes  across  Europe,  and  they  need  to  be  linked  theoretically  and 
empirically  to  the  study  of  more  ‘objective’  transition  patterns  exemplified  by  this 
project. 


A third  area  of  research  relates  to  policy  interventions.  A considerable  body  of 
research  focuses  on  the  impact  of  such  interventions,  particularly  on  State  training  or 
employment  schemes;  this  research  has  sparked  a good  deal  of  debate  about  the 
appropriate  way  to  measure  the  net  effect  of  such  intervention  programmes.  However, 
little  attention  has  been  given  to  developing  a broader  framework  for  considering 
policy  interventions  and  how  their  role  may  vary  across  different  contexts.  Do 
interventions  target  different  groups  of  young  people  in  different  countries?  How  do 
they  relate  to  'regular'  employment  and  to  the  initial  educational  system?  Are  there 
'trade-offs'  between  different  forms  of  policy  intervention?  An  understanding  of  these 
relationships  is  crucial  to  the  development  of  appropriate  interventions  in  different 
institutional  contexts. 

Fourthly,  research  has  tended  to  focus  on  national  variation  in  the  school  to  work 
transition  process  with  the  consequent  neglect  of  an  exploration  of  regional  and/or 
local  differences.  However,  labour  market  conditions  are  likely  to  vary  from  one  area 
to  another  as  is  the  supply  of  (certain  types  of)  education  and  training.  The  extent  to 
which  such  variation  is  significant  is  also  likely  to  relate  to  national-level  differences, 
for  example,  in  the  extent  of  decentralisation  of  policy-making.  An  exploration  of 
regional  differences  would,  therefore,  represent  a significant  contribution  to  our 
understanding  of  the  transition  process. 

In  addition,  further  research  is  needed  on  the  impact  of  field  of  education  on  transition 
outcomes  and  on  the  interrelationship  between  initial  education/training  and  access  to 
life-long  learning. 

4.5.2  Data  issues 

The  experience  of  carrying  out  analyses  as  part  of  the  CATEWE  project  highlighted 
the  strengths  and  weaknesses  of  existing  data  on  transitions. 

Analysis  of  the  European  Community  Labour  Force  Survey  yields  considerable 
insight  into  the  process  of  labour  market  integration  across  the  entire  spectrum  of 
European  countries.  However,  the  absence  of  'flow'  data  and  problems  concerning  the 
comparability  of  education  and  training  variables  and  of  labour  market  status  were 


highlighted  (Couppie  and  Mansuy,  1999a;  1999b).  It  is  recommended  that  the  use  of 
the  ECLFS  for  transitions  research  could  be  enhanced  through:  the  provision  of  more 
extensive  documentation  on  the  database  (covering  national  procedures  for  variable 
generation,  greater  details  on  sampling  and  so  on),  more  extensive  access  to  the 
database  for  scientific  research  and  improving  the  comparability  of  measures  over 
time  (Gangl  and  Muller,  1999).  It  is  felt  that  the  recent  module  on  transitions  has  great 
potential  for  cross-national  research  on  transitions.  It  is  recommended  that  such  a 
module  be  included  in  the  Labour  Force  Survey  on  a regular  basis  in  the  future. 
Attention  should  also  be  given  to  collecting  more  detailed  information  on  the 
transition  sequence,  the  experience  of  multiple  statuses  (e.g.  education  and  work), 
participation  in  training  and  employment  schemes,  and  social  background  information 
(Raffe,  2000). 

Existing  national  transition  surveys  were  found  to  vary  significantly  in  their  sample 
design,  purpose,  content  and  timing.  As  a result,  full  harmonisation  is  felt  to  be 
unrealistic  and  may  actually  be  counter  to  the  (national)  purpose  for  which  the 
surveys  are  designed  (Raffe,  2000).  However,  it  is  recommended  that  existing  and 
future  surveys  should  draw  on  an  agreed  set  of  criteria  to  achieve  partial 
harmonisation.  Surveys  should  aim  to  cover:  variables  which  reflect  equity  issues 
(such  as  social  background  and  ethnicity),  pathways  through  education,  training  and 
within  the  labour  market,  family/household  transitions,  measures  of  dual  status  (e.g. 
combining  education  and  work),  distinct  measures  of  inactivity,  time-related 
measures,  reference  transition  events  and  common  classifications  for  education, 
occupation  and  social  background  (Raffe,  2000).  In  terms  of  a European-wide  survey, 
it  is  considered  that  a prospective  age  cohort  design  would  have  considerable 
advantages  in  capturing  different  stages  of  the  transition  process  and  being  sensitive 
to  cross-national  differences  in  the  timing  and  sequencing  of  such  transitions.  While  a 
tailor-made  survey  along  these  lines  would  be  the  ideal,  the  planned  longitudinal 
element  of  the  PISA  survey  could  be  used  to  provide  comparable  information  on  the 
transition  from  school  to  the  labour  market. 


CHAPTER  FIVE:  DISSEMINATION  OF  PROJECT  RESULTS 


A number  of  complementary  strategies  were  adopted  within  the  life-time  of  the 
CATEWE  project.  Papers  based  on  the  conference  were  presented  at  a range  of 
conferences  at  both  national  and  European  levels.  These  included  the  European 
Research  Network  on  Transitions  in  Youth  annual  workshop,  the  European  Socio- 
Economic  Research  Conference  and  the  European  Sociological  Association  annual 
conference.  Furthermore,  the  CATEWE  team  organised  an  International  Workshop  on 
Comparative  Data  on  Education-to-Work  Transitions,  held  at  the  OECD  in  Paris  in 
June  2000  in  association  with  Network  B (on  transitions)  of  the  OECD’s  Educational 
Indicators  Project.  The  workshop  enabled  us  to  disseminate  our  work  to  policy- 
makers, statisticians  and  researchers  from  eighteen  countries  as  well  as  from  the 
European  Commission,  EUROSTAT,  CEDEFOP  and  the  OECD  itself.  It  also  gave  us 
the  opportunity  to  contribute  to  the  debate  on  proposals  for  future  data  harmonisation 
and  develop  links  with  interested  parties. 

A number  of  the  papers  presented  at  conferences  have  been  published  in  conference 
proceedings.  In  addition,  articles  on  the  project  findings  have  been  published  in  a 
number  of  policy-related  journals  and  CATEWE  contributions  will  be  part  of  a 
forthcoming  volume  on  Key  Data  on  Vocational  Training  along  with  a CEDEDOP 
publication  on  research  on  vocational  training.  The  CATEWE  working  papers  have 
also  been  made  available  on  our  web-site  (http://www.mzes.uni- 
mannheim.de/proiekte/catewe/).  Details  of  conference  papers  and  publications  are 
given  in  Table  5.1  below. 

At  present,  a book  proposal  on  the  Labour  Force  Survey  analyses  is  being  circulated 
to  relevant  publishers.  It  is  intended  to  submit  papers  on  the  School  Leaver  Survey 
analyses  to  scientific  journals.  We  also  intend  to  use  existing  links  among  project 
partners  to  engage  in  national-level  dissemination  to  policy-makers  and  other 
interested  parties. 


114 


121 


Table  5.1:  Papers  and  publications  by  the  CATE  WE  project  team 


Author(s) 

Title  of  paper  and  conference/publication 

1.  Publications 

Biggart,  Andy  and 
Raffe,  David  (2000) 

’Educational  backgrounds  and  transition  outcomes  in  four  European 
countries:  A comparison  of  school  leavers’  destinations  in  France, 
Ireland,  the  Netherlands  and  Scotland',  pp.l  13-131  in  T.  Hammer  (ed) 
Transitions  and  Mobility  in  the  Labour  Market:  Proceedings: 
Workshop , September  2-5  1999,  Oslo,  European  Network  on 
Transitions  in  Youth , Oslo:  NOVA. 

Brauns,  Hildegard, 
Gangl,  Markus  and 
Scherer,  Stefan  i 
(1998) 

'Education  and  Unemployment:  Patterns  of  Labour  Market  Entry  in 
France,  the  United  Kingdom  and  West  Germany',  pp.  287-312  in 
David  Raffe,  Rolf  van  der  Velden  and  Patrick  Werquin  (eds.), 
Education,  the  Labour  Market  and  Transitions  in  Youth:  Cross- 
National  Perspectives.  Proceedings  of  the  1998  European  Workshop 
of  the  European  Research  Network  on  Youth.  Edinburgh:  Centre  for 
Educational  Sociology. 

Couppie  T.,  Mansuy 
M.  (1998) 

'The  characteristics  of  youth  emplyment  in  Europe:  a typology  based 
on  the  Labour  Force  Surveys',  pp.  1-13  in  David  Raffe,  Rolf  van  der 
Velden  and  Patrick  Werquin  (eds.),  Education,  the  Labour  Market 
and  Transitions  in  Youth:  Cross-National  Perspectives.  Proceedings 
of  the  1998  European  Workshop  of  the  European  Research  Network 
on  Youth.  Edinburgh:  CES. 

Couppie  T.,  Mansuy 
M.  (2000) 

'The  situation  of  Young  labour-market  entrants  in  Europe',  Cereq, 
Training  and  employment,  n°39,  April-June. 

Couppie  T.,  Mansuy 
M.  (2000) 

'La  place  des  debutants  sur  les  marches  du  travail  Europeens',  Cereq, 
Bref  n°  154,  May. 

Gangl,  Markus  (2000) 

'European  Perspectives  on  Labour  Market  Entry:  A Matter  of 
Occupationalised  versus  Flexible  Arrangements  in  Labour 
Markets?',  pp.  417-452  in  Torild  Hammer  (ed.),  Transitions  and 
Mobility  in  the  Youth  Labour  Market.  Oslo:  NOVA. 

Gangl,  Markus, 
Hannan,  Damian, 
Raffe,  David  and 
Smyth,  Emer  (1998) 

'CATEWE  - A Comparative  Analysis  of  Transitions  from  Education 
to  Work  in  Europe’,  EURODATA  Newsletter , 8 (Autumn),  11-15. 

Hannan,  Damian  F.  et 
al.  (forthcoming) 

'The  effects  of  national  institutional  differences  in  education/training  to 
work  transitions  in  Europe:  a comparative  research  project  (CATEWE) 
under  the  TSER  programme',  in  Second  Report  on  Vocational 
Research  in  Europe,  CEDEFOP. 

Iannelli,  Cristina 
(2001) 

The  Effect  of  Schools  on  Young  People 's  Transitions:  A Comparative 
Perspective.  CES  Briefing.  University  of  Edinburgh,  forthcoming. 

Mansuy,  Michele, 
Couppie,  Thomas, 
Fetsi,  Anastasia, 
Scatoli,  Carlo 
Mooney,  Paul, 
van  den  Brande, 
Godelieve 

Key  Data  on  Vocational  Training  in  Europe:  The  Transition 
between  Education  and  Working  Life,  Volume  3,  forthcoming  spring 
2001,  European  Commission. 

Muller,  Walter  (1999) 

'Wandel  in  der  Bildungslandschaft  Europas'  [Changes  in  European 
educational  systems.!,  Gegenwartskunde,  1999:  Sonderband  1 1: 

Author  (s) 

Title  of  paper  and  conference/publication 

Deutschland  im  Wandel  - Sozialstrukturelle  Analysen,  pp.  337-356 

Muller,  Walter  (2000) 

'Bildung  in  Europa.  Zwischen  Tradition  und  Innovation' 

[Education  in  Europe.  Between  tradition  and  innovation.],  FORUM  - 
Forschung  Universitat  Mannheim,  2000:  pp.  34-41 

Muller,  Walter 

'Education  and  Labour  Market  Outcomes:  Commonality  or 
Divergence?'  In  Max  Haller  (ed.)  The  Making  of  the  European 
Union.  Contributions  of  the  Social  Sciences.  Berlin:  Springer,  2001, 
p 287-308. 

Muller,  Walter, 
Markus  Gangl  and 
Stefani  Scherer 
(forthcoming) 

'Ubergangsstrukturen  zwischen  Bildung  und  Beschaftigung'  [The 
structure  of  transitions  from  school  to  work.],  Forthcoming  in 
Matthias  Wingens  and  Reinhold  Sackmann  (eds.),  Bildung  und 
Beruf.  Weinheim  / Munich:  Juventa,  2001 . 

Muller,  Walter  and 
Wolbers,  Maarten 
(2000) 

'Educational  attainment  of  young  people  in  the  European  Union:  cross- 
country variation  of  trends  over  time',  pp.  263-302  in  Torild  Hammer 
(ed.),  Transitions  and  Mobility  in  the  Youth  Labour  Market.  Oslo: 
NOVA. 

Raffe,  David  (1999) 

'A  comparative  analysis  of  transitions  from  education  to  work  in 
Europe',  pp.  B/4-B/7  in  The  European  Socio-Economic  Research 
Conference:  Conference  Reader,  Brussels  28-30  April  1999,  European 
Commission. 

Raffe,  David  (2001) 

'The  social  construction  of  cross-national  transition  research:  The 
Network  on  Transitions  in  Youth',  (in  French  translation)  in  Education 
et  Societes,  thematic  issue  on  Entre  education  et  travail:  les  acteurs  de 
l 'insertion,  forthcoming. 

Smyth,  Emer  and 
McCoy,  Selina  (2000) 

'Differentation  among  school  leavers  in  the  transition  process:  a 
comparative  analysis',  pp.  303-341  in  Torild  Hammer  (ed.), 
Transitions  and  Mobility  in  the  Youth  Labour  Market.  Oslo:  NOVA. 

Velden,  Rolf  van  der, 
& Wolbers,  Maarten 
(2000). 

'De  integrate  van  school verlaters  op  de  arbeidsmarkt  in  de  Europese 
Unie:  de  rol  van  institutionele  factoren',  in  F.  Holderbeke,  R. 
Wielers  & N.  van  den  Heuvel  (red .),  De  transitionele  arbeidsmarkt. 
Contouren  van  een  actief  arbeidsmarktbeleid.  The  Hague:  Elsevier 
Bedrijfsinformatie  (in  press). 

2.  Working  papers  and 
reports 

Brauns,  Hildegard, 
Gangl,  Markus  and 
Scherer,  Stefani 
(1999) 

Education  and  Unemployment:  Patterns  of  Labour  Market  Entry  in 
France,  the  United  Kingdom  and  West  Germany.  MZES  Working 
Paper  No. 6.  Mannheim:  MZES. 

CATE  WE  (1999) 

Construction  of  a Comparative  Database  of  School  Leavers' 
Surveys:  A Report,  Report  to  the  European  Commission. 

CATEWE  (2000) 

A Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe  - Based  on  National  School  Leavers'  Surveys,  Report  to  the 
European  Commission. 

Gangl,  Markus  (2000) 

European  Perspectives  on  Labour  Market  Entry:  A Matter  of 
Institutional  Linkages  between  Training  Systems  and  Labour 
Markets?  MZES  Working  Paper  24.  Mannheim  : MZES 

Gangl,  Markus  (2000) 

Education  and  Labour  Market  Entry  across  Europe  : The  Impact  of 
Institutional  Arrangements  in  Training  Systems  and  Labour 
Markets.  MZES  Working  Paper  25.  Mannheim  : MZES 

Gangl,  Markus  (2000) 

Changing  Labour  Markets  and  Early  Career  Outcomes  : Labour 
Market  Entry  in  Europe  over  the  Past  Decade.  MZES  Working 
Paper  26.  Mannheim  : MZES 

Author(s) 

Title  of  paper  and  conference/publication 

Hannan  et  al.  (1999) 

A Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe , Volumes  1 and  2.  Dublin:  ESRI  Working  Paper  No.  118. 

Hannan,  Damian  F., 
Werquin,  Patrick 
(1999) 

Education  and  Labour  Market  Change:  The  Dynamics  of  Education  to 
Work  Transitions  in  Europe.  Report  to  the  European  Commission. 

Iannelli,  Cristina  and 
Soro-Bonmati, 
Asuncion  (2001, 
forthcoming) 

The  Transition  from  School  to  Work  in  Southern  Europe:  The  Cases 
of  Italy  and  Spain.  Mannheim:  MZES  Working  Paper. 

Muller,  Walter, 
Brauns,  Hildegard, 
Couppie,  Thomas, 
Gangl,  Markus, 
Mansuy,  Michele, 
Welters,  Riccardo  and 
Wolbers,  Maarten 
(1999) 

A Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe  - Based  on  the  European  Community  Labour  Force  Survey . 
Mannheim,  Maastricht,  Marseille:  Report  to  the  European 

Commission. 

Raffe,  David  (2000) 

Strategies  for  Collecting  Cross-National  Data  on  Education-to- 
Work  Transitions:  Recommendations  of  the  CATEWE  project . 
CATEWE  Deliverable  to  the  European  Commission. 

3.  Conference  papers 
(not  published  in 
proceedings) 

Biggart,  Andy  and 
Raffe,  David  (1999) 

Educational  Backgrounds  and  Transition  Outcomes  in  Four 
European  Countries.  Delivered  at  the  European  Research  Network 
on  Transitions  in  Youth:  Annual  Meeting,  Oslo,  Norway,  2 
September  1999. 

Biggart,  Andy  and 
Raffe,  David  (1999) 

Educational  Backgrounds  and  Transition  Outcomes  in  Four 
European  Countries . Delivered  at  the  European  Conference  on 
Educational  Research  1999,  Lahti,  Finland,  23  September  1999. 

Brannen,  Karen, 
Smyth,  Emer  (2000) 

Issues  in  constructing  a comparative  database  from  national 
transition  surveys.  Presentation  to  the  International  Workshop  on 
Comparative  Data  on  Education-to-Work  Transitions,  Paris,  June 
2000. 

Couppie  T.,  Mansuy 
M.  (2000) 

Using  the  Eurostat  LFS  as  a source  of  data  on  transition  in  Europe. 
OECD  Meeting,  Paris,  June  2000. 

Gangl,  Markus  (1998) 

CATEWE  - A Comparative  Analysis  of  Transitions  from  Education 
to  Work  in  Europe,  at  the  Joint  CEDEFOP/TSER  Conference  on 
Research  on  Vocational  Training,  June  1998, 

Thessaloniki/CEDEFOP. 

Gangl,  Markus  (1999) 

Education  and  Unemployment:  Patterns  of  Labour  Market  Entry  in 
France,  the  United  Kingdom,  and  West  Germany,  at  the  4th  ESA 
Conference  'Will  Europe  Work?',  Amsterdam,  August  1999  (Stefani 
Scherer  / Markus  Gangl) 

Gangl,  Markus  (1999) 

European  Perspectives  on  Labour  Market  Entry:  A Matter  of 
Occupationalized  versus  Flexible  Arrangements  in  Labour 
Markets?,  at  the  1999  Annual  Workshop  of  the  European  Research 
Network  on  Transitions  in  Youth,  Oslo,  September  1999 

Gangl,  Markus  (1999) 

Unemployment  Risks  in  Early  Labour  Market  Careers:  Results  from 
Labour  Force  Surveys  and  Implications  for  the  Use  of  Longitudinal 
Data,  at  the  SONAR- Workshop  of  the  Faculty  of  Economics  and 
Management,  University  of  Gent,  November  1999 

Gangl,  Markus  (1999) 

Education  and  Unemployment:  Patterns  of  Labour  Market  Entry  in 

Author(s) 

Title  of  paper  and  conference/publication 

France,  the  United  Kingdom,  and  West  Germany,  at  the  MZES 
Workshop  on  'Education  and  the  Labour  Market',  January  1999 
(Hildegard  Brauns  / Stefani  Scherer  / Markus  Gangl) 

Gangl,  Markus  (2000) 

Using  LFS  Data  for  Cross-National  Research:  Promises,  Examples, 
and  Problems,  at  the  International  Workshop  on  Comparative  Data 
on  Education-to-Work  Transitions,  OECD,  Paris,  June  2000  (Walter 
Muller  / Markus  Gangl) 

Gangl,  Markus  (2000) 

Using  LFS  Data  for  Cross-National  Research  on  School-to-Work 
Transitions:  Examples  from  the  CATE  WE  Project,  at  EUROSTAT, 
Unit  E-3:  Education,  health  and  other  social  fields,  Luxembourg, 
October  2000  (Walter  Muller  / Markus  Gangl) 

Gangl,  Markus  (2000) 

Education  and  Labour  Market  Entry  across  EU  Countries:  Looking 
for  the  Effects  of  Institutional  Contexts , at  the  Annual  Meeting  of  the 
MZES  Advisory  Board,  February  2000 

Gangl,  Markus  (2000) 

Ausbildung  und  Berufseinstieg  in  Europa:  Bildungssysteme, 

Institutionelle  Effekte,  und  okonomischer  Nontext  als  Erklarung  der 
Landerunterschiede  im  Ubergangsprozefi  zwischen  Schule  und 
Beruf?  [Education  and  Labour  Market  Entry  in  Europe:  Educational 
Systems,  Institutional  Effects,  and  Macroeconomic  Context  as 
Determinants  of  Cross-National  Differences  in  Transition  Patterns.], 
at  the  ABA  Research  Seminar,  June  2000 

Gangl,  Markus  (2000) 

Simulationstechniken  in  der  komparativen  Sozialforschung 
[Simulation  analysis  in  comparative  social  research],  at  the  30. 
KongreB  der  Deutschen  Gesellschaft  fur  Soziologie,  Cologne,  26-29 
September  2000 

Grelet  Y.,  Mansuy  M., 
Thomas  G.  (2000) 

How  does  the  transition  process  lead  to  financial  sufficiency  or 
not?,  European  Research  Network  on  Transitions  in  Youth  2000 
meeting,  Antwerpen  (Belgium) 

Hannan,  Damian  F., 
Werquin,  Patrick 
(1999) 

Education  and  Labour  Market  Change:  The  Dynamics  of  Education  to 
Work  Transitions  in  Europe,  Presentation  to  the  TSER  Socio- 
Economic  Research  Conference,  Brussels,  28-30  April. 

Hartkamp,  Jannes,  and 
Rutjes,  Hans  (2000) 

A Route  to  Skills:  A Comparative  Analysis  of  the  Position  of 
Apprenticeship  in  Transition  Systems  in  France,  Ireland,  the 
Netherlands  and  Scotland.  Delivered  at  the  European  Research 
Network  on  Transitions  in  Youth:  Annual  Meeting,  Antwerp,  7-9 
September  2000. 

Iannelli,  Cristina 
(2000) 

School  Effect  on  Youth  Transitions  in  Ireland,  the  Netherlands  and 
Scotland.  Delivered  at  the  European  Research  Network  on 
Transitions  in  Youth:  Annual  Meeting,  Antwerp,  7-9  September 
2000. 

Iannelli,  Cristina  and 
Raffe,  David  (2000) 

Vocational  Upper  Secondary  Education  and  the  Transition  from 
School.  Delivered  at  the  Euroconference  on  Educational 
Differentiation  in  European  Societies:  Causes  and  Consequences, 
Giens,  16-21  September  2000. 

Mansuy  M.,  Schroder 
L.  (2000) 

Young  immigrants  on  the  labour  Market  in  France  and  Sweden, 
European  Research  Network  on  Transitions  in  Youth  2000  meeting, 
Antwerpen  (Belgium) 

Raffe,  David  (1999) 

The  Gender  Dimension  in  Socio-Economic  Research:  Comparative 
Analysis  of  Transitions  from  Education  to  Work  in  Europe. 
Delivered  at  the  European  Socio-Economic  Research  Conference, 
Brussels,  April  1999.  

Raffe,  David  (1999) 

Transitions  from  Education  to  Work  in  Europe.  Seminar,  University 
of  Essex,  29  November  1999. 

Author(s) 

Title  of  paper  and  conference/publication 

Raffe,  David  (2000) 

Strategies  for  Collecting  Cross-National  Data  on  Education-to- 
Work  Transitions:  Recommendations  of  the  CATE  WE  project. 
Delivered  at  the  International  Workshop  on  Comparative  Data  on 
Education-to-Work  Transitions,  Paris,  France,  21  June  2000. 

Raffe,  David  (2000) 

Strategies  for  Collecting  Cross-National  Data  on  Education-to- 
Work  Transitions:  Recommendations  of  the  CATEWE  project . 
Delivered  at  the  European  Research  Network  on  Transitions  in 
Youth:  Annual  Meeting,  Antwerp,  Belgium,  9 September  2000. 

Raffe,  David  (2000) 

The  role  of  vocational  training  and  education  in  the  combat  against 
youth  unemployment.  Delivered  at  the  International  Expert  Forum 
on  Winning  the  Future:  Strategies  against  Youth  Unemployment, 
EXPO,  Hannover,  Germany,  25  October  2000. 

Schroder,  Lena  (2000) 

Unemployment  among  Immigrants  and  Ethnic  Minorities:  The 
Potential  of  Active  Labour  Market  Programmes.  Delivered  at  the 
International  Expert  Forum  on  Winning  the  Future:  Strategies 
against  Youth  Unemployment,  EXPO,  Hannover,  Germany,  October 
2000. 

Smyth,  Emer  (2000) 

Gender  differentiation  in  education  and  early  labour  market 
transitions:  a comparative  analysis.  Delivered  at  the  European 
Research  Network  on  Transitions  in  Youth:  Annual  Meeting, 
Antwerp,  7-9  September  2000. 

Wolbers,  Maarten 
(1999) 

Educational  attainment  of  young  people  in  the  European  Union: 
Cross-country  variation  of  trends  over  time.  Paper  presented  at  T3e 
Onderwijssociologische  Conference',  Amsterdam,  10-11  October 
1999. 

Wolbers,  Maarten 
(2000) 

The  integration  of  young  people  into  the  labour  market  within  the 
European  Union  : The  role  of  institutional  settings.  Paper  presented 
at  '2e  Vlaams-Nederlandse  Arbeidsmarktcongres',  Brussels,  2 Maart 
2000. 

Wolbers,  Maarten 
(2000) 

The  integration  of  young  people  into  the  labour  market  within  the 
European  Union:  The  role  of  institutional  settings.  Paper  presented 
at  8th  annual  meeting  of  the  European  Research  Network  on 
Transitions  in  Youth,  Antwerp,  Belgium,  7-10  September  2000. 

CHAPTER  SIX:  REFERENCES 


CA  TEWE  Working  Papers 

Biggart,  Andy  and  Raffe,  David  (2000).  'Educational  backgrounds  and  transition 
outcomes  in  four  European  countries:  A comparison  of  school  leavers’ 
destinations  in  France,  Ireland,  the  Netherlands  and  Scotland',  pp.  113-131  in  T. 
Hammer  (ed.)  Transitions  and  Mobility  in  the  Labour  Market:  Proceedings: 
Workshop,  September  2-5  1999,  Oslo,  European  Network  on  Transitions  in 
Youth,  Oslo:  NOVA. 

Brannen,  Karen  and  Smyth,  Emer  (2000).  'Issues  in  constructing  a comparative 
database  from  national  transition  surveys',  paper  to  the  International 
Workshop  on  Comparative  Data  on  Education- to- Work  Transitions,  Paris. 

Brauns,  Hildegard,  Markus  Gangl,  and  Scherer,  Stefani  (1999).  Education  and 
unemployment:  Patterns  of  labour  market  entry  in  France,  the  United 
Kingdom,  and  West  Germany.  MZES  Working  Paper  6.  Mannheim:  MZES. 

CATEWE  (1999).  Construction  of  a Comparative  Database  of  School  Leavers' 
Surveys:  A Report,  Report  to  the  European  Commission. 

Couppie,  Thomas  and  Mansuy,  Michele  (1999).  'Appendix  A:  Methodological 
Problems',  in  Muller,  W.  et  al.,  A Comparative  Analysis  of  Transitions  from 
Education  to  Work  in  Europe  - Based  on  the  European  Community  Labour 
Force  Survey,  Report  to  the  European  Commission. 

Couppie,  Thomas  and  Mansuy,  Michele  (2000a).  The  Position  of  New  Entrants  on 
European  Labour  Markets.  CATEWE  Working  Paper.  Marseille:  mimeo. 

Couppie,  Thomas  and  Mansuy,  Michele  (2000b).  New  Entrants  and  experienced 
workers  on  European  Labour  Markets.  CATEWE  Working  Paper.  Marseille: 
mimeo. 

Gangl,  Markus  (1999).  'European  Perspectives  on  Labour  Market  Entry:  A Matter  of 
Occupationalised  versus  Flexible  Arrangements  in  Labour  Markets?',  in 
Muller,  W.  et  al.,  A Comparative  Analysis  of  Transitions  from  Education  to 
Work  in  Europe  - Based  on  the  European  Community  Labour  Force  Survey, 
Report  to  the  European  Commission. 

Gangl,  Markus  (2000a).  European  Perspectives  on  Labour  Market  Entry:  A Matter  of 
Institutional  Linkages  between  Training  Systems  and  Labour  Markets?  MZES 
Working  Paper  24.  Mannheim:  MZES. 

Gangl,  Markus  (2000b).  Education  and  Labour  Market  Entry  across  Europe:  the 
Impact  of  Institutional  Arrangements  in  Training  Systems  and  Labour 
Markets.  MZES  Working  Paper  25.  Mannheim:  MZES. 


120 


127 


Gangl,  Markus  (2000c).  Changing  Labour  Markets  and  Early  Career  Outcomes: 
Labour  Market  Entry  in  Europe  over  the  Past  Decade.  MZES  Working  Paper 
26.  Mannheim:  MZES. 

Gangl,  Markus  and  Muller,  Walter  (1999).  'Appendix  C:  Transition  research  using  the 
European  Community  Labour  Force  Survey',  in  Muller,  W.  et  al.,  A 
Comparative  Analysis  of  Transitions  from  Education  to  Work  in  Europe  - 
Based  on  the  European  Community  Labour  Force  Survey,  Report  to  the 
European  Commission. 

Grelet,  Yvette,  Mansuy,  Michele,  Thomas,  Gwenaelle  (2000a).  'Transition  from 
school  to  work  and  early  labour  force  history',  in  A Comparative  Analysis  of 
Transitions  from  Education  to  Work  in  Europe  - Based  on  National  School 
Leavers'  Surveys,  Report  to  the  European  Commission. 

Grelet,  Yvette,  Mansuy,  Michele,  Thomas,  Gwenaelle  (2000b).  The  transition 
process:  towards  exclusion  or  financial  sufficiency,  a French-lrish 

comparison.  CATEWE  Working  Paper,  Marseille:  mimeo. 

Hannan,  Damian  F.  et  al.  (1999).  A Comparative  Analysis  of  Transitions  from 
Education  to  Work  in  Europe,  Volumes  1 and  2.  Dublin:  ESRI  Working  Paper 
No.  118. 

Hartkamp,  Jannes  and  Rutjes,  Hans  (2000a).  'A  route  to  skills:  a comparative  analysis 
of  the  position  of  apprenticeship  in  transition  systems  in  France,  Ireland,  the 
Netherlands  and  Scotland',  in  A Comparative  Analysis  of  Transitions  from 
Education  to  Work  in  Europe  - Based  on  National  School  Leavers'  Surveys, 
Report  to  the  European  Commission. 

Hartkamp,  Jannes  and  Rutjes,  Hans  (2000b).  Apprenticeship  in  Ireland,  the 
Netherlands  and  Scotland:  comparison  of  trends  1979-1997.  CATEWE 
Working  Paper,  Amsterdam:  mimeo. 

Iannelli,  Cristina  (2000).  School  effects  on  youth  transitions  in  Ireland,  Scotland  and 
the  Netherlands.  CATEWE  Working  Paper,  Edinburgh:  mimeo. 

Iannelli,  C.  and  Raffe,  D.  (2000).  'Vocational  upper-secondary  education  and  the 
transition  from  school  to  work',  in  A Comparative  Analysis  of  Transitions  from 
Education  to  Work  in  Europe  - Based  on  National  School  Leavers'  Surveys, 
Report  to  the  European  Commission. 

Iannelli,  Cristina,  and  Soro-Bonmati,  Asuncion  (2000).  The  Transition  from  School  to 
Work  in  Southern  Europe:  The  Cases  of  Italy  and  Spain.  CATEWE  Working 
Paper.  Mannheim:  mimeo. 

Mansuy,  Michele  and  Schroder,  Lena  (2000).  Immigrant  youth  in  the  labour  market 
in  France  and  Sweden.  CATEWE  Working  Paper,  Marseille/Stockholm: 
mimeo. 


O 

ERIC 


121 


128 


McCoy,  Selina  (2000a).  'Relative  labour  market  disadvantage  amongst  the  least 
qualified  in  Ireland,  Scotland,  the  Netherlands,  France  and  Sweden',  in  A 
Comparative  Analysis  of  Transitions  from  Education  to  Work  in  Europe  - 
Based  on  National  School  Leavers’  Surveys,  Report  to  the  European 
Commission. 

McCoy,  Selina  (2000b).  Relative  labour  market  disadvantage  among  the  least 
qualified:  Ireland,  the  Netherlands  and  Scotland,  1979-1997.  CATEWE 
Working  Paper.  Dublin:  mimeo. 

Muller,  Walter  et  al.  (1999).  A Comparative  Analysis  of  Transitions  from  Education 
to  Work  in  Europe  - Based  on  the  European  Community  Labour  Force  Survey, 
Report  to  the  European  Commission. 

Muller,  Walter  and  Wolbers,  Maarten  (1999).  'Educational  attainment  of  young 
people  in  the  European  Union:  cross-country  variation  of  trends  over  time',  in 
Muller,  W.  et  al.,  A Comparative  Analysis  of  Transitions  from  Education  to 
Work  in  Europe  - Based  on  the  European  Community  Labour  Force  Survey, 
Report  to  the  European  Commission. 

Raffe,  David  (2000).  Strategies  for  collecting  cross-national  data  on  education-to- 
work  transitions:  recommendations  of  the  CATEWE  project.  Report  to  the 
European  Commission. 

Schroder,  Lena  (2000).  'The  role  of  youth  programme  in  the  transition  from  school  to 
work',  in  A Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe  - Based  on  National  School  Leavers'  Surveys,  Report  to  the  European 
Commission. 

Smyth,  Emer  (2000a).  'Gender  differentiation  in  education  and  transition  outcomes', 
in  A Comparative  Analysis  of  Transitions  from  Education  to  Work  in  Europe  - 
Based  on  National  School  Leavers'  Surveys,  Report  to  the  European 
Commission. 

Smyth,  Emer  (2000b).  Gender  differentiation  in  education  and  early  labour  market 
outcomes  over  time:  a comparative  analysis.  CATEWE  Working  Paper, 
Dublin:  mimeo. 

Smyth,  Emer  and  McCoy,  Selina  (2000).  'Differentiation  among  school  leavers  in  the 
transition  process:  a comparative  analysis',  in  T.  Hammer  (ed.)  Transitions  and 
Mobility  in  the  Labour  Market:  Proceedings:  Workshop,  September  2-5  1999, 
Oslo,  European  Network  on  Transitions  in  Youth,  Oslo:  NOVA. 

Van  der  Velden,  Rolf,  and  Wolbers,  Maarten  (2000).  The  integration  of  young  people 
into  the  labour  market  within  the  European  Union:  the  role  of  institutional 
settings.  CATEWE  Working  Paper.  Maastricht:  mimeo. 

Welters,  R.  and  Wolbers,  Maarten  (1999).  'Learning  and  working:  double  statuses  in 
youth  transitions  within  the  European  Union',  in  Muller,  W.  et  al.,  A 
Comparative  Analysis  of  Transitions  from  Education  to  Work  in  Europe  - 

122 

1 9 0 


ERIC 


Based  on  the  European  Community  Labour  Force  Survey,  Report  to  the 
European  Commission. 

Wolbers,  Maarten  (2000).  Learning  and  working:  Double  statuses  in  youth  transitions 
within  the  European  Union.  CATEWE  Working  Paper,  Maastricht:  mimeo. 

Other  references 

Allmendinger,  Jutta  (1989).  'Educational  systems  and  labour  market  outcomes’, 
European  Sociological  Review,  5:  231-250. 

Blanchflower,  David  D.,  and  Freeman,  Richard  B.  (2000).  'The  Declining  Economic 
Status  of  Young  Workers  in  OECD  Countries',  in  David  D.  Blanchflower  and 
Richard  B Freeman  (Eds.),  Youth  Employment  and  Joblessness  in  Advanced 
Countries.  Chicago  and  London:  University  of  Chicago  Press. 

Bynner,  John  and  Chisholm,  Lynne  (1998).  'Comparative  youth  transition  research: 
methods,  meanings  and  research  relations',  European  Sociological  Review,  14, 
pp. 131-150. 

CEDEFOP  (1994).  The  Determinants  of  Transitions  in  Youth.  Papers  from  the 
conference  organized  by  the  ESF  Network  on  Transitions  in  Youth, 
CEDEFOP  and  GRET.  Berlin:  CEDEFOP. 

Cereq  (1997).  Passage  des  jeunes  gens  du  systeme  scolaire  a la  vie  active. 
Verification  des  enquetes  et  des  informations  existantes  (Enquetes  Forces  de 
Travail  et  ECHP).  Rapport  Final.  Marseille:  Cereq. 

ESRI/  Combat  Poverty  Agency  (1998).  Transitions  in  Youth:  Combating  Exclusion. 
Proceedings  of  the  Fourth  European  Workshop.  Dublin:  ESRI/CPA. 

Evans,  Karen  (1999).  Comparative  successes  or  failures?  Some  methodological  issues 
in  conducting  international  comparative  research  in  PCET.  Paper  to  British 
Educational  Research  Association  (University  of  Surrey,  School  of 
Educational  Studies). 

Evans,  Karen  and  Heinz,  Walter  R.  (eds.)  (1996).  Becoming  Adults  in  England  and 
Germany.  London:  Anglo-German  Foundation. 

Garonna,  P.  and  Ryan,  P.  (1991)  'The  regulation  and  deregulation  of  youth  economic 
activity',  in  P.  Ryan,  P.  Garonna.  and  R.  Edwards  (Eds)  The  Problem  of  Youth, 
London:  Macmillan. 

Hammer,  Torild  (1997).  'History  Dependence  in  Youth  Unemployment',  European 
Sociological  Review  13,  17-33. 

T.  Hammer  (ed.)  (2000).  Transitions  and  Mobility  in  the  Labour  Market:  Proceedings: 
Workshop,  September  2-5  1999,  Oslo,  European  Network  on  Transitions  in 
Youth,  Oslo:  NOVA. 


Hannan,  Damian  F.,  McCabe,  Breda,  McCoy,  Selina  (1998).  Trading  Qualifications 
for  Jobs:  Overeducation  and  the  Irish  Youth  Labour  Market.  Dublin:  Oak  Tree 
Press/  ESRJ. 

Hannan,  Damian  F.,  Smyth,  Emer,  et  al.,  (1998).  Education,  Vocational  Training  and 
Labour  Market  Transitions  amongst  Lower  Level  Leavers  in  Four  European 
Countries,  Report  to  the  European  Commission. 

Hannan,  Damian,  Raffe,  David,  and  Smyth,  Emer  (1997).  'Cross-National  research  on 
school  to  work  transitions:  An  analytic  framework',  in  P.  Werquin,  R.  Breen 
and  & Jordi  Planas  (Eds),  Youth  Transitions  in  Europe:  Theories  and 
Evidence,  Documents  Seminaires  No.  120,  Marseille:  CEREQ. 

Heinz,  Walter  R.  (ed.)  (1999).  From  Education  to  Work:  Cross-National 
Perspectives.  Cambridge:  Cambridge  University  Press. 

Kohn,  M.  (1987).  'Cross-national  research  as  an  analytic  strategy',  American 
Sociological  Review,  52,  pp.7 13-731. 

Marsden,  David  (1986).  The  End  of  Economic  Man?  Brighton:  Wheatsheaf. 

Martin,  Chris  and  Raffe,  David  (1998).  'Lower  level  leaving  and  desertion  - their  level 
and  social  distribution  in  four  European  countries',  in  Hannan,  Damian  F.  et 
al.,  Education,  Vocational  Training  and  Labour  Market  Transitions  amongst 
Lower  Level  Leavers  in  Four  European  Countries,  Report  to  the  European 
Commission. 

Maurice,  M.,  Sellier,  F.  and  Silvestre,  J-J.  (1986)  The  Social  Foundations  of 
Industrial  Power:  A comparison  of  France  and  Germany.  Cambridge,  Mass: 
MIT  Press. 

McIntosh,  Steven  and  Steedman,  Hilary  (1999).  Low  Skills:  A Problem  for  Europe. 
Final  Report  to  DGXII  of  the  European  Commission  on  the  NEWSKILLS 
programme  of  research,  London:  LSE. 

Muller,  Walter  and  Shavit,  Yossi  (1998).  'The  institutional  embeddedness  of  the 
stratification  process',  in  Yossi  Shavit  and  Walter  Muller  (eds.),  From  School 
to  Work,  Oxford:  Clarendon  Press. 

OECD  (1997).  Education  at  a Glance.  OECD  Indicators  1997.  Paris:  OECD. 

OECD  (2000).  From  Initial  Education  to  Working  Life:  Making  Transitions  Work. 
Paris,  OECD. 

0yen,  E.  (ed.)  (1990).  Comparative  Methodology.  London:  Sage. 

Raffe,  David  (2001).  'The  social  construction  of  cross-national  transition  research: 
The  network  on  Transitions  in  Youth',  CES,  University  of  Edinburgh. 
Forthcoming  in  Education  et  Societes : special  issue  Entre  education  et  travail: 
les  acteurs  de  l ’insertion. 


ERIC 


124 


1 *3  7 


David  Raffe,  Rolf  van  der  Velden  and  Patrick  Werquin  (eds.)  (1999),  Education,  the 
Labour  Market  and  Transitions  in  Youth:  Cross-National  Perspectives. 
Proceedings  of  the  1998  European  Workshop  of  the  European  Research 
Network  on  Youth.  Edinburgh:  Centre  for  Educational  Sociology. 

Ragin,  C.  and  Becker,  H.  (Eds)  (1992).  What  is  a case?  Cambridge:  Cambridge 
University  Press. 

Roberts,  K.,  Clark,  S.C.,  and  Wallace,  C.  (1994).  'Flexibility  and  individualisation:  a 
comparison  of  transitions  into  employment  in  England  and  Germany', 
Sociology,  Vol.  28,  No.  1. 

Shavit,  Yossi  and  Blossfeld,  Hans  Peter  (eds.)(1993).  Persistent  Inequality:  Changing 
Educational  Attainment  in  Thirteen  Countries.  Boulder:  Westview  Press. 

Shavit,  Yossi  and  Muller,  Walter  (2000).  'Vocational  secondary  education:  where 
diversion  and  where  safety-net?',  European  Societies,  Vol.  2,  No.  1,  pp. 29-50. 


A COMPARATIVE  ANALYSIS  OF  TRANSITIONS 
FROM  EDUCATION  TO  WORK  IN  EUROPE  (CATEWE) 


ANNEX  TO  THE  FINAL  REPORT 


CATEWE  PROJECT  DELIVERABLES 


Status  of  deliverables 


Deliverable 

Status 

Content  deliverables 

1 . Paper  outlining  a conceptual 

framework  for  transitions  in  Europe 

Completed 

2.  International  workshop 

Joint  workshop  with  the  OECD  Thematic 
Review  team  in  June  1998 

3.  Cross-national  database  on  school 
leavers 

Completed  for  internal  project  use; 
expanded  to  construct  three  separate 
datasets 

4.  Paper  on  school  to  work  transitions  in 
a European  context  (using  Eurostat 
Labour  Force  Survey  data) 

Completed 

5.  Paper  on  school  to  work  transitions 
(using  school  leavers'  survey  data) 

Completed 

6.  International  workshop  on  data 
harmonisation 

Held  in  June  2000 

7.  Proposals  for  a European- wide  school 
leavers'  survey 

Completed 

Contractual  deliverables 

1 . Six-monthly  progress  reports  (M6, 
M12,  Ml 8,  M24,  M30) 

Completed 

2.  Twelve-monthly  cost  statements 
(M12.M24) 

Completed 

Working  papers  which  have  been  written  or  substantially  revised  since  the  submission 
of  these  deliverables  are  presented  in  the  remainder  of  this  volume. 


January  2001 


Prepared  as  part  of  the  TSER  project: 

Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe 


Education  and  Unemployment: 

Patterns  of  Labour  Market  Entry  in  France,  the 
United  Kingdom  and  West  Germany 


Hildegard  Brauns 
Markus  Gangl 
Stefani  Scherer 


Mannheim  Centre  for  European  Social  Research  (MZES) 

University  of  Mannheim 

P.O.  Box  10  34  62 

D-68131  Mannheim 

Germany 

E-mail:  Markus.Gangl@mzes.uni-mannheim.de 


WORKING  PAPERS 


Abstract 


Over  the  last  two  decades,  youth  unemployment  emerged  as  one  of  the  major  problems  of  many 
contemporary  European  societies.  As  educational  achievement  is  regularly  argued  to  prevent  labour 
market  exclusion,  this  paper  explores  the  educational  stratification  of  unemployment  in  the  early 
labour  market  career  and  its  institutional  embeddedness  in  specific  education  and  employment 
systems.  For  the  sake  of  comparative  analysis,  the  paper  investigates  youth  unemployment  in  France, 
the  United  Kingdom  and  West  Germany  as  these  three  countries  differ  greatly  in  terms  of  major 
institutional  characteristics  of  their  educational  systems  and  labour  markets.  The  analyses  use 
microdata  from  national  Labour  Force  Surveys  of  the  mid-1980s  and  the  mid-1990s  allowing  an 
assessment  of  recent  trends  as  well  as  comparative  analysis.  Methodologically,  we  rely  on  single- 
stage  and  sequential  logit  models  to  estimate  the  effects  of  individual  educational  achievement  on 
unemployment  risks.  As  a result,  we  are  able  to  present  evidence  of  a sharp  distinction  between  the 
educational  stratification  observed  in  Germany  on  the  one  hand  and  France  and  the  United  Kingdom 
on  the  other.  In  Germany,  labour  market  entry  is  found  to  be  quite  smooth  and  immediate  for 
vocationally  qualified  leavers,  while  extensive  periods  of  searching  for  a first  job  is  confined  almost* 
exclusively  to  the  least  qualified.  After  initial  employment  has  been  found,  education  plays  a negligible 
role  in  the  risk  of  unemployment  which  is  tied  more  to  aspects  of  employment  positions.  In  France  and 
Britain,  in  contrast,  the  match  between  qualifications  and  jobs  is  less  clear-cut.  Rather,  the  level  of 
education  provides  advantages  in  terms  of  reduced  time  searching  for  employment  and  lower  job 
instability,  although  differentiation  is  much  less  pronounced.  In  addition,  education  effects  maintain  a 
positive  impact  on  job  stability  even  controlling  for  positional  characteristics,  suggesting  a more 
gradual  match  between  qualifications  and  attainment.  Results  are  found  to  be  stable  for  both  time 
periods,  indicating  idiosyncratic  rather  than  secular  changes  in  the  educational  stratification  of  youth 
unemployment  over  the  last  decade. 


137 


1.1 


1 Introduction 


In  many  European  societies,  the  rising  incidence  of  youth  unemployment  has  been  one  of  the  major 
social  problems  since  the  1970s.  A significant  number  of  young  people  experience  unemployment 
after  completing  their  education,  are  likely  to  be  unemployed  again  years  later  and  to  face  extended 
periods  of  social  marginalisation  during  their  early  careers.  While  investment  in  education  has 
traditionally  been  seen  as  the  best  means  for  young  people  to  secure  a good  job,  the  formal  education 
and  training  system  is  presently  subject  to  severe  criticism  for  its  deficiencies  in  endowing  school- 
leavers  with  marketable  skills.  Against  this  background,  it  is  a central  concern  of  both  policy  makers 
and  social  science  researchers  to  explore  the  reasons  why  young  people  have  these  difficulties  in 
getting  and  keeping  a job,  the  way  in  which  educational  credentials  bring  them  “in  the  door"  (Bills 
1988),  and  the  way  in  which  distinctive  arrangements  in  the  institutionalisation  of  education  make  a 
difference  in  securing  “employability”  for  school-leavers. 

The  present  paper  analyses  the  incidence  of  youth  unemployment  with  reference  to  individuals’- 
educational  achievement,  and  explores  how  the  educational  stratification  of  unemployment  varies 
across  countries  and  across  time  in  the  course  of  educational  expansion.  The  countries  to  be 
compared  are  West-Germany,  France  and  the  United  Kingdom.  The  three  countries  have  all 
experienced  massive  educational  expansion  over  recent  years,  but  still  differ  considerably  in  the 
institutionalisation  of  their  education  and  labour  market  systems.  While  the  cross-national  perspective 
will  shed  light  on  the  impact  of  the  specific  institutional  environment  on  the  labour  market  value  of 
single  educational  credentials,  the  historical  perspective  is  supposed  to  give  some  evidence  on  the 
extent  to  which  this  value  depends  on  the  educational  resources  of  all  other  job  candidates  in  the 
market,  or  in  other  words,  on  the  relative  scarcity  of  the  achieved  credential. 

Despite  a substantial  amount  of  comparative  work  on  individual  returns  to  education  on  the  one  hand, 
and  of  research  on  unemployment  on  the  other,  the  two  lines  of  research  are  hardly  connected.  Most 
comparative  work  on  individual  education  returns  focuses  on  work  related  outcomes  such  as  income, 
occupational  prestige,  social  class  position  or  access  to  so-called  ‘good’  versus  ‘bad’  jobs.  However,  in 
times  of  slack  labour  markets,  it  is  essential  to  also  investigate  the  way  in  which  educational 
credentials  help  people  'get  in  the  door’.  Micro-level  comparative  research  on  unemployment,  by 
contrast,  is  sparse.  The  few  existing  studies  are  almost  exclusively  restricted  to  German-American 
comparisons  and  typically  focus  on  individual  unemployment  dynamics,  the  issue  of  duration 


Financial  support  from  the  Deutsche  Forschungsgemeinschaft  and  the  TSER  programme  of  the  European 
Commission  is  gratefully  acknowledged.  We  thank  Josef  Bruderl  from  the  University  of  Mannheim  and  Heike 
Solga  from  the  Max-Planck  Institute  for  Human  Development  in  Berlin  for  helpful  comments.  We  also  thank 
Wolfgang  Fischer  for  his  extensive  help  in  editing.  The  paper  was  presented  at  the  ASA  meeting  in  Chicago  in 
August  1999  and  the  ESA  meeting  in  Amsterdam  in  August  1999;  a previous  version  was  presented  at  the  annual 
meeting  of  the  European  network  on  Transition  in  Youth  in  Edinburgh  in  September  1998  and  the  MZES 
Workshop  'Bildung  und  Arbeitsmarkt'  in  Mannheim  in  January  1999. 


ERIC 


1.2 


dependence  and  on  the  role  of  benefit  disincentives.  The  present  paper  aims  to  address  to  this 
research  gap.  It  will  be  guided  by  the  following  main  questions: 

Which  role  does  education  play  with  regard  to  young  people’s  risk  of  unemployment?  To  what  extent 
does  the  importance  of  educational  credentials  in  unemployment  risk  depend  on  the  career  stage  of 
individuals? 

Do  we  find  evidence  of  distinct  national  linkages  between  education  and  unemployment  risks?  If  so,  in 
what  way  does  the  institutional  set-up  in  the  three  countries  shape  the  educational  stratification  of 
unemployment  risks? 

How  does  the  increasing  supply  of  school-leavers  with  higher  education  over  the  course  of 
educational  expansion  affect  the  distribution  of  unemployment  risks?  Is  there  a common  secular  trend 
across  nations? 

The  remainder  of  the  paper  proceeds  as  follows:  the  next  section  presents  theoretical  perspectives 
and  central  research  questions  that  will  be  used  to  guide  our  analysis.  Section  3 describes  data  and' 
empirical  procedures.  The  analyses  draw  on  individual  data  from  national  labour  force  surveys  carried 
out  in  the  mid-1980s  and  mid-1990s.1  Empirical  results  are  the  subject  of  sections  4 to  6.  Section  4 
gives  descriptive  results  on  overall  levels  and  types  of  unemployment  at  the  two  survey  points  in  the 
three  countries.  In  sections  5 and  6,  the  risk  of  unemployment  is  related  to  educational  achievement. 
In  a first  step,  the  rate  of  unemployment  experienced  by  each  educational  group  and  the  respective 
distribution  of  relative  advantages  and  disadvantages  according  to  education  are  compared  across 
time  and  country.  In  a second  step,  we  go  a little  further  into  the  details  of  unemployment  risks  in  early 
labour  market  careers,  starting  from  the  idea  that  the  benefits/risks  attached  to  formal  education  per 
se  depend  on  the  career  stage  of  the  individual  at  risk  of  unemployment.  It  will  be  argued  that  formal 
education  is  a highly  important  and  differentiating  resource  for  school-leavers  who  are  searching  for 
their  first  job  immediately  after  leaving  school,  while  it  is  a still  significant,  but  less  stratifying  asset 
once  young  people  have  succeeded  in  finding  employment,  and  thus  transformed  their  educational 
resources  into  positional  resources  that  more  or  less  protect  them  against  unemployment.  However, 
the  extent  to  which  formal  education  matters  more  in  the  hiring  stage  than  in  the  stage  after  job 
placement  has  taken  place  is  expected  to  vary  between  countries  due  to  institutional  differences  in  the 
educational  and  employment  systems.  The  paper  concludes  with  a summary  of  the  empirical  results. 


' Surveys  from  the  mid-1980s  and  mid-1990s  are  chosen  because  both  periods  depict  periods  of  peak 
unemployment  in  each  country.  The  more  tense  the  labour  market  situation,  the  stronger  the  competition  for  jobs 
and  the  more  demanded  background  characteristics  escalate  due  to  selective  hiring  (Thurow  1975).  We  should 
then  be  able  to  identify  strong  educational  effects  and  potential  differences  in  the  patterns  across  countries. 
Furthermore,  the  similarity  in  labour  market  situation  across  time  allows  us  to  hold  business  cycle  effects  constant 
and,  thereby,  to  capture  the  effects  of  educational  expansion. 


1.3 


2 Theoretical  Perspectives 

Both  economic  and  sociological  theories  acknowledge  a strong  relationship  between  education  and 
unemployment,  although  a variety  of  (conflicting)  mechanisms  are  specified  for  explaining  the  linkage 
(Mincer  1994;  Spence  1973,  1981;  Kettunen  1994).  Signalling  and  screening  models  argue  that 
employers  hire,  place  and  promote  workers  on  the  basis  of  imperfect  information  about  their  true 
productivity  (Stigler  1961;  Arrow  1973;  Spence  1973  among  others).  Educational  credentials  are 
interpreted  as  a filter  that  serves  primarily  as  a measure  of  performance  ability  for  the  employers, 
important  to  the  extent  that  other  reliable  indicators  are  lacking.  Thurow’s  (1975)  "queuing"  version  of 
screening  theory  contends  that  employers  rank  order  the  desirability  of  job  candidates  according  to 
their  trainability  for  given  jobs.  Credentials  indicate  which  job  candidates  are  most  and  which  are  least 
likely  to  be  trained  for  a given  job.  Individuals  are  distributed  across  job  opportunities  on  the  basis  of 
their  relative  position  in  the  labour  queue.  In  times  of  labour  surplus,  school  leavers  who  are  placed  at 
the  bottom  of  the  labour  queue  will  be  unemployed. 

Educational  achievement,  however,  has  many  dimensions  that  may  signal  competence  or. 
incompetence  for  a given  job.  As  a regularity,  we  assume  that 

HI : the  level  of  general  education  reached  and  vocational  specificity  of  the  degree  earned  make  a 

difference  for  one’s  chances  of  being  offered  a job. 

First,  employers  look  for  information  about  job  candidates’  general  cognitive  capacities  to  learn  new 
skills  and  to  adapt  to  new  technical  environments.  Therefore,  they  rely  to  some  extent  on  the  sorting 
done  by  the  educational  system,  since  educational  processes  already  involve  long  periods  of 
screening  and  subsequent  selection  (Jencks  1972).  Second,  employers  are  seeking  employees  who 
already  have  some  expertise  in  the  tasks  at  hand,  in  order  to  minimise  the  costs  of  training. 
Achievement  of  vocational  skills  should  then  also  be  valued  on  the  labour  market. 

Employers  do  more  than  just  educational  screening,  however.  Gender,  date  of  leaving  school  and 
prior  work  experience,  where  this  information  is  available,  are  other  important  sources  of  information 
about  employees  in  labour  markets.  The  crucial  question  is  how  much  weight  is  placed  on  education 
as  a screening  device  and  to  what  extent  does  this  depend  on  the  circumstances  under  which  young 
people  are  at  risk  of  unemployment.  According  to  the  reasoning  of  Thurow  (1978), 

H2:  educational  achievement  should  be  a more  differentiating  resource  for  initial  hiring  into  the 

labour  force  than  after  first  job  placement  has  taken  place,  that  is  for  workers  with  prior  job 
experience. 


0 

ERIC 


The  relative  weight  of  education  is  thus  assumed  to  be  dependent  on  the  individual’s  career  stage. 
Upon  leaving  school,  when  job-related  performance  records  are  either  not  available  or  are  only 
scarcely  available  to  employers,  they  tend  to  rely  primarily  on  the  candidates'  performance  in  the 
educational  and  training  system.  Once  the  candidate  has  been  employed  and  has  succeeded  in 
transforming  his  educational  resources  into  positional  resources,  which  more  or  less  protect  him 


1.4 


against  unemployment,  the  role  of  education  as  a performance  indicator  should  to  some  extent  be 
superseded  by  direct  performance  records.  Nonetheless,  we  expect  educational  achievement  to 
continue  to  have  an  effect  after  entry  into  employment.  First,  plausible  information  about  individuals' 
productive  capabilities  unfolds  only  slowly  with  time.  Therefore,  in  the  early  labour  market  career, 
education  is  continuously  used  as  an  important  screening  device  in  employers’  personnel  decisions. 
Second,  due  to  environmental  pressures  employers  also  reward  credentials  as  a matter  of  firm  policy 
(Spilerman  1986).  Third,  credentials  are  rewarded  for  their  social  value  (Spilerman  and  Lunde  1991; 
Collins  1979;  Bowles  and  Gintis  1976).  Credentials  indicate  conventional  standards  of  sociability,  the 
ease  of  adapting  to  new  tasks  and  the  capacity  to  internalise  organisational  rules  and  firm  culture 
which  makes  their  holders  "promotable”. 

Thurow  (1975:  93ff)  also  argues,  however,  that  the  concrete  shape  of  the  labour  queue,  i.e.  the  rank 
order  of  school-leavers  and  the  “distance"  between  them,  is  determined  by  two  factors:  employer 
preferences  for  certain  credentials  and  the  distribution  or  dispersion  of  educational  achievement 
among  the  job  candidates.  Referring  to  the  first  factor,  larger  differences  in  the  benefit/risk  pertaining 
to  the  various  educational  credentials  may  result  if  employers  have  a pronounced  preference  for 
certain  credentials.  This  “preference"  for  certain  credentials  over  others,  that  is  for  certain  aspects  of 
educational  achievement,  is  significantly  shaped  by  institutional  features  of  the  national  education 
system,  the  labour  market  and  their  interrelationship  (see  Breen  et  al.  1995;  Muller/Shavit  1998; 
Brauns  et  al.  1999;  Kerckhoff  forthcoming,  1996;  Hannan  et  al.  1997).  Overview  1 gives  a description 
of  the  educational  systems  in  the  three  countries  and  their  link  to  the  labour  market.2 

As  a consequence  of  substantial  cross-national  differences  in  the  institutionalisation  of  educational 
and  labour  market  systems,  employers’  “preferences”  for  certain  facets  of  educational  achievement 
should  vary  between  the  three  countries. 

In  Germany,  occupational  labour  markets  prevail  in  which  jobs  are  clearly  defined  in  terms  of  content 
and  occupational  skill  requirements.  Therefore,  we  expect  German  employers  to  primarily  reward 
occupational  significance  of  the  achieved  certificate  when  selecting  among  job  candidates. 

H3:  compared  to  school-leavers  with  general  education  only,  vocationally  qualified  school-leavers 

should  profit  from  substantial  advantages  with  regard  to  labour  market  entry. 

Vocationally  qualified  school-leavers  gain  a number  of  advantages  compared  to  school-leavers  with 
general  education  only.  First,  given  that  apprenticeship-training  is  by  far  the  most  widespread  form  of 
vocational  education  in  Germany,  vocationally  qualified  school-leavers  are  highly  skilled  in  an 
occupation,  already  socialised  into  working  life  and  into  the  organisational  culture  of  the  company. 


2 The  classifications  of  the  three  labour  market  types  are  only  ideal-type  descriptions.  Since  the  analysis  proceeds 
only  at  the  national  level,  it  intends  to  express  the  predominance  of  a certain  type  of  labour  market  structure  in  a 
given  national  setting.  All  three  labour  markets  are  indeed  made  up  of  occupational  as  well  as  firm-internal  labour 
market  structures  (see  Kalleberg/Berg  1987). 


ERiC 


1.5 


Overview  1 


Germany  is  typically  associated  with  the  predominance  of  occupationally  structured  labour  markets  where  jobs 
are  standardised  nationwide,  clearly  defined  in  terms  of  content  and  occupational  skill  requirements  (see  Maurice 
et  al.  1982;  Eyraud/Marsden/Silvestre  1990;  Marsden  1990;  Marsden/Ryan  1991a,  1995).  This  labour  market  type 
is  said  to  be  closely  linked  to  the  strong  emphasis  on  vocational  training  in  the  German  education  system,  which 
confers  highly  standardised,  occupationally-specific  qualifications.  The  large  majority  of  German  school-leavers, 
from  all  levels  of  general  education,  have  completed  vocational  training  which  predominantly  takes  place  within 
the  dual,  that  is  the  apprenticeship  training  system.  The  dual  system  provides  a high  degree  of  homogeneity 
throughout  the  system.  It  is  organised  along  the  lines  of  occupations,  almost  400  in  the  skilled  trades  and 
administration,  industry,  services,  agriculture,  health  etc.  Compared  to  vocational  education  systems  in  most  other 
countries,  the  German  system  is  not  hierarchically  stratified,  in  that  hierarchical  qualification  tracks  are  offered 
which  require  specific  attainment  in  the  general  education  system  for  admission.  Although  pupils'  choices  of 
occupations  to  be  trained  in  are  not  regulated  by  educational  policy,  one’s  level  of  general  education  attained 
largely  determines  the  chances  and  choices  made.  Overall,  the  concept  of  a "Beruf”  (an  occupation)  is  a central 
principle  that  regulates  training  content,  qualification  standards  and  examinations,  in  particular  in  the  dual  system 
but  also  in  higher  education.  Higher  education  is  also  shown  to  be  less  stratified,  horizontally  and  vertically,  but 
also  less  expanded  than  in  other  countries  (see  OECD  1998).  Academic  training  is  traditionally  offered  by  the 
universities,  involving  at  least  four  years  of  study.  Shorter,  less  academic  studies,  although  on  a highly  theoretical 
level,  are  provided  by  the  “Fachhochschulen”. 

France  is  typically  associated  with  the  prevalence  of  internal  labour  markets.  In  contrast  to  occupational  labour 
markets,  job  design  and  skill  requirements  are  highly  firm-specific.  In  internal  labour  markets,  recruitment  into  af 
limited  number  of  entry  points  occurs,  found  at  the  bottom  of  firm-specific  job  ladders.  On  these  ladders 
individuals  may  gradually  progress  from  one  job  to  the  other  which  entails  the  progressive  development  of  skills 
and  knowledge.  The  weak  emphasis  in  the  French  education  system  on  vocational  education,  and  on 
apprenticeship  training  in  particular,  is  considered  an  important  indication  of  this  labour  market  type.  Compared  to 
Germany,  the  French  education  system  traditionally  has  a much  stronger  emphasis  on  general  than  on  vocational 
education.  Over  the  last  two  decades,  however,  vocational  education  facilities  have  been  expanded  and 
modernised.  Yet,  vocational  training  is  still  combined  with  a preference  for  the  schoolroom  over  the  workplace  for 
vocational  preparation.  General  and  vocational  education  are  embedded  within  a single  integrated  system  which 
is  highly  stratified.  Below  the  level  of  tertiary  education,  qualifications  are  available  on  four  distinctive  levels,  each 
of  which  is  differentiated  into  various  tracks.  Vocational  qualifications  alone  are  offered  on  three  levels  and  in 
three  different  branches,  the  latter  being  differentiated  into  a number  of  tracks  (see  Brauns/Steinmann  1999; 
Brauns  forthcoming).  From  the  institutional  interweaving  of  vocational  and  general  education  and  the  selection 
regime  in  the  French  education  system  (see  Brauns  1998)  it  follows  that  achievement  of  vocational  qualifications 
is  highly  related  to  (non-)achievement  in  the  general  school  system.  Likewise,  tertiary  education  is  highly 
stratified,  both  horizontally  and  - on  three  levels  - vertically.  Long-term  studies  are  available  at  the  universities, 
and  on  a smaller  scale,  at  the  elitist  "Grandes  Ecoles".  Short-term  studies  in  some  academic  areas  are  awarded 
by  the  universities,  and  on  a much  smaller  scale,  by  the  elitist  preparatory  courses  for  the  ‘Grandes  Ecoles”.  More 
practical  courses  are  taught  by  a number  of  other  institutions  that  have  been  designed  for  the  training  of  highly 
specialised  technicians,  nurses,  kindergarten  teachers  etc. 

There  are  conflicting  views  on  the  British  labour  market.  Disregarding  the  relatively  small  segment  of  the  skilled 
trades,  where  occupational  labour  markets  traditionally  prevail,  and  given  the  preference  for  "generalists”  rather 
than  "specialists"  in  the  British  labour  market,  one  would  tend  to  classify  the  labour  market  as  one  where  internal 
rather  than  occupational  labour  market  structures  prevail  (see  Sorge  1983;  Lane  1992;  Marsden/Ryan  1991b). 
Although  Great  Britain  shares  the  tradition  of  apprenticeships  with  Germany,  there  are  substantial  commonalities 
with  France:  first,  apprenticeships  have  not  attained  the  same  prestige  as  in  Germany  nor  the  same  wide  diffusion 
across  all  economic  sectors.  They  are  mainly  confined  to  the  crafts  and  have  been  declining  steadily  since  the 
1960s.  Second,  vocational  education  has  in  general  always  been  of  secondary  importance.  As  in  France,  it  does 
not  carry  any  status  in  the  wider  society  and  does  not  attract  the  more  able  and  motivated  youngsters.  The  British 
system  has  a stronger  emphasis  on  general  education  and  on  producing  "generalists”  rather  than  "specialists” 
trained  in  a specific  occupation.  In  the  early  1990s,  however,  the  British  started  an  initiative  to  modernise 
vocational  training  which  had  previously  been  largely  unregulated  and  unstandardised.  Vocational  qualifications 
are  offered  on  different  levels  and  in  two  different  frameworks:  broad-based  General  National  Vocational 
Qualifications  (GNVQs)  and  job-specific  National  Vocational  Qualifications  (NVQs).  Higher  education  involves  a 
"binary  system”  with  the  universities  and  the  polytechnics  combined  with  other  technical  colleges. 


Second,  apprenticeships  serve  a screening  function.  With  apprentices,  employers  have  drawn  on  a 
two-  to  three-year  period  of  intensive  face-to-face  screening  and  can  make  their  hiring  decisions  with 
considerable  confidence  about  job  applicants’  productivity.  Third,  the  German  dual  system  operates 
according  to  market  rules  insofar  as  training  capacities  depend  on  companies’  willingness  to  offer 
training  places.  As  a consequence,  young  people  who  were  offered  an  apprenticeship  have  a good 
chance  of  remaining  with  the  company  afterwards.  Fourth,  dual  system  training  is  highly  standardised 
nationwide.  Vocational  qualification  certificates  confer  highly  reliable  information  about  the 
apprentices’  skills  even  for  employers  who  hire  other  companies’  apprentices. 

In  France  and  the  UK,  firm  internal  labour  markets  rather  than  occupational  labour  markets  are 
predominant.  Compared  to  Germany,  this  implies  a limited  importance  of  apprenticeship-training,  a 
lower  standardisation  of  jobs  across  firms  and  a less  “institutionalised”  linking  of  school-leavers’  labour 
market  opportunities  to  their  educational  achievement.  In  consequence,  occupational  skills  are 
supposed  to  be  relatively  unimportant  for  recruitment  into  first  job.  Instead,  employers  should  look  for 
signals  indicating  job  candidates’  cognitive  capabilities  to  be  trained  for  the  firm-specific  tasks. 

H4  In  France  and  the  UK,  labour  market  entry  is  expected  to  be  less  selective  but  also  less 
“smooth”  than  in  Germany.  Rather  than  valuing  vocational  qualifications,  employers  in  these 
countries  should  screen  job  candidates  according  to  their  level  of  education  attained. 

In  internal  labour  markets,  school-leavers  should  not  find  themselves  excluded  from  being  hired  simply 
on  the  strength  of  their  educational  (non)achievement.  Since  entrants  have  not  yet  developed  the 
skills  that  are  required  for  the  job,  entry  is  typically  confined  to  the  “bottom"  of  internal  career  ladders, 
often  implying  a “qualification-inadequate”  occupational  position  and  a temporary  employment 
contract.  The  benefits  relating  to  specific  credentials  typically  arise  later,  in  the  light  of  the  internal 
regulation  of  careers,  when  decisions  about  promotion  and  continued  employment  are  made.  This 
should  be  the  case  particularly  in  France.  France  is  a country  with  a strong  credentialist  tradition 
where  formal  education  plays  an  important  role  in  legitimising  peoples’  social  standing  and  careers 
(Brauns  1998).  This  credentialist  orientation  has  been  reinforced  by  state  employment  policy  by 
means  of  a national  ’’qualification-grid"  aimed  at  establishing  "correspondence”  between  people’s 
educational  performance  and  their  employment  situation. 

Occupational  labour  markets,  by  contrast,  are  characterised  by  a “structured”  transition  from  training 
into  employment,  namely  a regulated,  standardised  career  line.  Due  to  the  institutionalised  link 
between  types  of  educational  pathways  and  occupational  “entitlement”,  vocationally  qualified  school- 
leavers  who  are  offered  a job  in  their  occupation  are  almost  sure  to  secure  a "good”  job  match. 
Therefore,  once  people  are  hired  into  the  closed  system,  they  convert  their  educational  resources  into 
certain  positional,  labour  market  related  resources.  Chances  of  career  advancement  and  of  security  of 
employment  are  now  closely  tied  to  these  positional  resources  rather  than  to  formal  education.  Hence, 

H5:  net  of  the  employment  position  that  an  individual  occupies  in  the  labour  market,  continued 

employment  in  the  German  labour  market  is  expected  to  be  far  less  dependent  on  the 


ERIC 


1.7 


employee’s  educational  resources  than  is  the  case  in  the  British  and  especially  the  French 
labour  market. 

The  second  factor  that  Thurow  (1975:  93ff)  emphasises  as  shaping  the  labour  queue  is  the  distribution 
or  dispersion  of  educational  achievement  among  the  job  candidates.  All  three  countries  have 
experienced  a massive  expansion  of  education  over  the  last  decades  resulting  in  a substantial 
increase  in  the  average  educational  attainment  of  job  candidates.  What  does  this  development  mean 
for  different  educational  groups  with  regard  to  their  exposure  to  unemployment?  The  job  competition 
model  assumes  that  employers’  skill  requirements  are  responsive  to  changes  in  the  relative  supply  of 
different  skill  groups.  As  a result  of  educational  expansion,  employers  will  attempt  to  acquire  higher 
quality  labour  than  in  earlier  periods.  A greater  proportion  of  higher  educated  groups  will  then  be 
filtered  into  lower-paying  and  unstable  jobs  than  in  earlier  times,  which  were  previously  allocated  to 
lower  qualified  school-leavers.  In  a chain  reaction,  the  latter  are  then  increasingly  pushed  out  of  the 
labour  market.  Assuming  that  there  are  no  offsetting  changes  in  the  demand  side  of  the  market3, 

H6:  in  times  of  labour  surplus,  educational  expansion  should  foster  increasing  of  labour  market 

entry  difficulties  especially  among  the  least  qualified. 

Following  the  argument  made  above  on  the  role  of  institutional  context,  however,  it  seems  plausible  to 
expect  the  consequences  of  educational  expansion  for  particular  educational  groups  to  vary  across 
countries.  We  have  no  specific  hypotheses  on  country  differences  in  the  impact  of  educational 
expansion.  The  empirical  analysis  should  shed  some  light  on  this  issue. 


3 Data  and  Methodology 

The  empirical  analyses  draw  on  micro  data  from  national  Labour  Force  Surveys  (LFS).  Labour  Force 
Surveys  provide  detailed  information  about  the  current  employment  situation  and  details  about  the 
educational  achievements  as  well  as,  though  to  a limited  extent,  information  about  previous 
employment  situations.  They  have  many  advantages  in  comparative  research:  the  surveys  share 
central  methodological  practices,  a core  set  of  questions  and  measurement  set-ups.  This  commonality 
makes  them  very  valuable  for  constructing  cross-nationally  and  historically  comparable  indicators.  In 
addition,  the  huge  sample  sizes  available  in  the  LFS  allow  for  precise  statistical  estimation  and 
detailed  analyses  of  different  educational  groups.  In  this  paper,  we  use  the  1994  and  1984  Enquete 


3 However,  the  actual  distribution  of  employment  chances  among  school-leavers  is  not  only  a function  of  the 
labour  queue,  supply-side  characteristics,  but  also  of  the  distribution  of  job  opportunities,  that  is  the  availability  of 
jobs  with  specific  skill  requirements.  In  other  words,  demand  side  factors  may  also  generate  a change  in 
educational  requirements  over  time.  In  this  paper  we  will  not  be  able  to  empirically  disentangle  the  many  factors 
that  may  create  changing  returns  to  single  credentials:  reform  of  the  institutional  set-up  of  the  educational  system, 
changes  in  the  skill  dispersion  among  cohorts  of  school-leavers,  demand  side  factors  due  to  technological 
developments  and  sectoral  change  etc.  The  aim  of  this  paper  is  to  capture  the  combined  outcome  of  all 
processes. 


Emploi  for  France,  the  1994  and  1984  Labour  Force  Survey  for  the  United  Kingdom,  and  the  1995  and 
1982  Mikrozensus  for  Germany.  The  German  samples  are  confined  to  the  West  German  population, 
given  the  specific  situation  in  the  former  GDR,  which  would  require  separate  investigation. 

This  paper  analyses  unemployment  risks  in  individuals’  early  labour  market  careers,  or  as  we  call  it,  in 
the  period  of  transition  from  school  to  work4.  The  primary  focus,  however,  is  not  on  explaining  the 
dynamics  of  unemployment,  but  on  the  impact  of  education  on  unemployment  risk  in  the  early  career5. 
In  our  definition,  the  transition  period  covers  up  to  seven  years  after  leaving  school.  We  include  those 
persons  who  left  the  education  and  training  system  within  the  seven  years  before  the  date  of  inquiry  in 
the  sample.  This  sample  selection  is  derived  from  information  on  the  year  of  leaving  full-time 
education  available  in  the  French  and  British  surveys6.  Since  the  German  surveys  do  not  offer  similar 
information,  the  date  of  completion  of  initial  education  and  training  was  calculated  synthetically.  Given 
the  stratified  nature  of  the  educational  system,  this  strategy  seems  reasonably  justified7.  As  a result  of 
this  restriction  and  after  excluding  individuals  not  in  the  labour  force  (for  definition  see  below),  in 
military  service  or  in  institutions,  we  have  total  sample  sizes  of  13,047  (France  1984),  11,767  (France 
1994),  15,527  (United  Kingdom  1984),  11,465  (United  Kingdom  1994),  42,684  (Germany  1982)8,  and 
19,494  cases  (Germany  1995). 

In  defining  the  labour  force,  we  basically  adopt  the  ILO  definition  with  slight  modifications9. 
Unemployment  is  measured  following  the  ILO  convention,  as  being  jobless  but  available  for  paid 
employment  within  two  weeks,  and  actively  searching  for  a new  job10.  Education  is  measured  by  the 
CASMIN  scale  developed  for  comparative  research  (Brauns/Steinmann  1999;  Muller/Shavit  1998; 
Konig/Muller/Luttinger  1988  for  the  original  conception).  We  apply  the  eight-category  version  of  the 
classification  that  is  shown  in  table  1.  The  scale  distinguishes  hierarchical  levels  of  attainment  and 
differentiates  between  ‘general  education’  and  ‘vocationally-orientated’  tracks.  Therefore,  the  CASMIN 
classification  allows  for  the  straightforward  representation  of  non-linearities  in  the  impact  of  education 
(see  Braun/Muller  1997). 


4 We  also  use.  the  terms  ‘youth  unemployment’  or  ‘unemployment  risks  in  the  early  career’  interchangeably 
throughout  the  paper. 

5 So,  this  perspective  relates  to  the  stratificational  power  of  education  in  the  different  countries  with  regard  to  the 
risk  of  unemployment  as  one  possible  labour  market  outcome  among  others. 

6 This  definition  explicitly  excludes  those  in  apprenticeships.  We  applied  an  additional  correction  for  those  who 
completed  some  work-based  apprenticeship  or  other  youth  training,  since  the  original  question  does  not  make 
any  reference  to  such  tracks.  The  correction  consists  of  adding  two  years  to  the  date  of  leaving  full-time  education 
in  the  case  of  leavers  from  apprenticeship  training  in  France,  and  of  leavers  from  apprenticeships  or  youth  training 
in  the  United  Kingdom. 

7 This  measure  was  calculated  by  assuming  9 years  of  schooling  for  the  completion  of  Hauptschule,  10  years  for 
Realschule,  and  13  years  for  Gymnasium.  To  this,  2 to  3 years  of  training  were  added  in  the  case  of  participation 
in  the  vocational  training  system  (dependent  upon  general  qualifications  being  present),  and  university  degrees 
have  been  taken  as  four  and  five  additional  years  of  education  respectively  (Fachhochschule  versus  traditional 
university  degree).  We  also  correct  for  young  males’  participation  in  compulsory  national  service. 

8 For  the  1980s  LFS  in  Germany  we  can  draw  on  a larger  sample  size,  namely  a 98%  sample.  To  take  into 
account  that  these  data  are  given  in  table  format,  a weighting  had  to  be  introduced  (Frenzel/Luttinger,  1990; 
Statistisches  Bundesamt,  1982). 

9 Individuals  participating  in  the  education  and  training  system  irrespective  of  their  detailed  status  are  not  regarded 
as  belonging  to  the  labour  force;  individuals  who  are  in  full-time  education  or  work-based  apprenticeship-type 
programmes  (apprenticeship  or  youth  training)  at  the  time  of  the  survey  are  thus  excluded  from  all  analyses. 

10  Individuals  who  are  presently  out  of  work,  but  are  to  take  up  a job  in  the  near  future  are  included  among  the 
unemployed  as  well,  according  to  the  ILO-definition. 


3 

ERIC 


1.9 


145 


Table  1:  The  CASMIN  scale  of  educational  qualifications 


Qualification  Description 

lab  This  is  the  social  minimum  of  education.  Namely,  the  minimal  level  that  individuals  are  expected 

to  have  obtained  in  a society.  It  generally  corresponds  to  the  level  of  compulsory  education 

1c  Basic  vocational  training  above  and  beyond  compulsory  schooling 

2b  Academic  or  general  tracks  at  the  secondary  intermediate  level 

2a  Advanced  vocational  training  or  secondary  programmes  in  which  general  intermediate  schooling 

is  combined  with  vocational  training 

2c  Full  maturity  certificates  (e.g.  the  Abitur,  Matriculation,  Baccalaur6at,  A-levels) 

2c  voc  Full  maturity  certificates  including  vocationally-specific  schooling  or  training  (e.g.  Baccalaur6at  de 

technicien) 

3a  Lower-level  tertiary  degrees,  generally  of  shorter  duration  and  with  a vocational  orientation  (e.g. 

technical  college  diplomas,  social  worker  or  non-university  teaching  certificates) 

3b  The  completion  of  a traditional,  academically-oriented  university  education 

Source:  adapted  from  Brauns/Steinmann  1997,  pp.  33-35,  and  Muller/Shavit  1998,  p.17 


In  the  empirical  analysis  we  adopt  different  perspectives  focusing  on  the  link  between  education  and 
unemployment.  Following  a first  description  of  the  phenomenon  in  each  of  the  three  countries,  we' 
explore  the  role  of  educational  achievement  on  overall  levels,  duration  and  reasons  for  unemployment. 
Cross-national  differences  and  changes  over  time  in  the  link  between  education  and  unemployment  in 
the  early  career  are  investigated  with  regard  to  absolute  rates  and  relative  risks.  Absolute  rates  of 
unemployment  within  single  education  groups  are  supposed  to  give  an  intuitive  understanding  of 
actual  risk  of  exposure  to  unemployment.  Odds  ratios  are  used  to  gain  a precise  understanding  of 
relative  unemployment  risks,  or  - put  differently  - inequalities  in  the  exposure  of  various  educational 
groups  to  unemployment.  Odds  ratios  express  the  competitive  (dis)advantage  in  terms  of 
unemployment  risk  that  holders  of  a specific  educational  credential  have  relative  to  others.  They  are 
derived  from  logistic  regression  models  which  are  applied  to  our  total  sample  populations. 

In  a complementary  step,  we  explore  the  role  of  education  in  more  detail  by  considering  the 
circumstances  under  which  individuals  are  at  risk  of  unemployment.  We  differentiate  between  two 
sequential  situations:  the  initial  search  period  and  (in)stability  in  the  early  career  after  a first  entry  into 
employment  has  taken  place. 

The  first  stage  focuses  on  the  risk  of  failing  to  enter  the  labour  market  at  all.  It  is  the  hiring  stage 
immediately  after  leaving  school  where  the  individual  searches  for  his  first  job.  At  this  stage,  since  job 
candidates  have  no  previous  job  experience,  employers  are  supposed  to  rely  on  job  candidates’ 
educational  achievement  as  a major  performance  indicator.  The  second  stage  is  the  situation  after 
successful  labour  market  entry,  in  which  the  person  is  at  risk  of  losing  the  job  (being  laid  off  in  most 
circumstances)  and  not  being  re-employed.  At  this  stage,  the  candidate  has  obtained  initial  work 
experience  which  the  employer  can  draw  on  when  deciding  about  displacement  or  re-employment.  At 
the  same  time,  the  candidate  has  succeeded  in  transforming  his  or  her  education  into  positional 
resources  which  more  or  less  protects  him  or  her  against  labour  market  exclusion.  While  the  first  stage 
is  a pure  recruitment  situation  the  picture  is  obviously  more  complex  with  regard  to  the  second  step.  At 


1.10140 


this  career  stage,  unemployment  risks  arise  from  different  processes,  namely  firing,  quitting  and  not 
being  re-employed. 

More  precisely,  we  distinguish  the  two  stages  based  on  the  information  on  whether  an  individual  has 
ever  had  a paid  job.  The  first  stage  includes  all  individuals  and  contrasts  those  who  are  presently 
unemployed  without  having  taken  up  any  kind  of  employment  before  against  those  who  have 
succeeded  in  finding  a job,  regardless  of  whether  they  are  employed  or  unemployed  at  the  time  of 
inquiry.  Information  about  individuals’  date  of  completing  their  initial  education  and  training  allows  us  to 
control  for  the  time  since  leaving  the  ETS.  Thus,  in  the  first  step,  we  model  the  probability  of  having 
found  employment,  or  in  other  words,  the  probability  of  still  being  in  the  state  of  initial  search 
unemployment  at  the  time  of  inquiry,  subject  to  individuals’  educational  achievement  and  the  time 
since  leaving  the  ETS.  In  the  second  stage,  we  focus  on  the  probability  of  unemployment  after  a first 
transformation  of  education  resources  into  job  resources  has  taken  place.  We  therefore  include  only 
those  respondents  who  have  already  taken  up  paid  work  at  the  time  of  inquiry.  This  allows  us  to 
assess  the  (in)stability  of  the  early  career,  depending  on  individuals’  educational  achievement  and 
characteristics  of  their  job.  Given  the  cross-sectional  character  of  LFS  data,  even  with  this  partition  we 
are  only  able  to  estimate  the  equilibrium  outcomes  in  each  stage,  not  the  underlying  actual  processes 
taking  place.  It  seems  worth  emphasising  that  LFS  data  does  not  provide  longitudinal  information 
about  the  whole  period  after  leaving  the  ETS11.  Thus,  what  we  measure  in  the  first  stage  is  not  the 
duration  to  first  employment  but  the  likelihood  of  having  been  in  paid  employment  at  the  date  of 
inquiry.  With  the  second  stage  we  assess  the  probability  of  not  finding  re-employment  at  the  time  of 
inquiry,  for  those  who  lost  or  quit  their  previous  job,  and  not  the  general  incidence  of  unemployment. 

Two  separate  logit  models  are  applied  to  estimate  the  effects  of  education  at  either  stage.  While  the 
first  step  controls  for  waiting  time  after  completion  of  initial  education  and  training,  the  second  step 
takes  individuals’  labour  market  position  and  contract  situation  into  consideration.  The  labour  market 
position  distinguishes  between  jobs  in  the  unskilled,  skilled  and  the  professional  segment  or  in  self- 
employment,  drawing  on  information  about  the  preceding  job  for  those  who  are  currently  unemployed. 
The  classification  is  based  on  EGP  class  position  (Erikson/Goldthorpe,  1992; 
Brauns/Haun/Steinmann,  1997)  and  information  about  the  employment  sector.  Details  are  given  in  the 
tables  in  Appendix  A1.  The  contract  situation  considers  temporary  contracts  in  contrast  to  non-limited 
employment  situations. 


4 Unemployment  at  Labour  Market  Entry:  an  Overview 

The  resurgence  of  persistent  unemployment  has  been  a common  European  experience  during  the 
past  two  decades.  In  many  respects,  young  people  have  been  severely  affected  by  the  employment 


11  This  also  means  that  we  include  people  at  different  stages  in  their  early  career  in  our  sample,  and  that  we  are 
not  able  to  disentangle  cohort,  period  and  age  effects.  But  with  the  rather  narrow  observation  window  of  up  to 
seven  years  since  leaving  the  ETS  it  seems  reasonable  to  expect  age  effects  to  dominate. 


1.11 


147 


crises  of  the  1980s  and  1990s:  youth  unemployment  rates  are  regularly  found  to  be  substantially 
higher  than  those  of  the  adult  labour  force  (Layard  et  al.  1991;  OECD  1 996),  the  smoothness  of  labour 
market  integration  is  often  claimed  to  be  strongly  affected  by  prevailing  structural  conditions  (e.g. 
Blossfeld  1989)  and  whether  such  difficulties  at  entering  the  labour  market  inflict  permanent  ‘scars’  on 
subsequent  employment  histories  is  a constant  issue  of  intensive  research  and  concern.  However, 
many  observers  also  argue  that  cross-national  differences  in  youth  unemployment  are  much  more 
pronounced  than  among  the  adult  work-force  (cf.  Layard  et  al.  1991),  which  could  actually  provide  an 
indication  of  the  effects  of  institutional  arrangements  of  market  entry  operating.  What  can  we  tell  about 
such  differences  in  the  level  and  structure  of  youth  unemployment  between  France,  the  United 
Kingdom  and  West  Germany  on  a descriptive  level?  As  is  apparent  from  the  results  presented  in  table 
2 below,  the  overall  picture  is,  in  many  respects,  one  of  significant  difference  between  the  countries  in 
terms  of  major  aspects  of  unemployment  patterns. 

Considering  the  results  in  more  detail,  the  risk  of  unemployment  is  smallest  in  Germany,  where 
unemployment  rates  were  at  9%  in  1982  and  6%  in  1995.  In  contrast,  unemployment  rates  were 
highest  in  France  with  some  23%  unemployed  at  both  points  in  time.  The  United  Kingdom  occupies  an 
intermediate  position:  unemployment  rates  were  20%  in  1984  and  18%  in  1994.  Youth  unemployment 
has  thus  been  a fairly  substantial  problem  in  both  France  and  the  United  Kingdom  throughout  the  last 
decade,  while  labour  market  entry  has  been  considerably  smoother  in  West  Germany.  Additionally, 
unemployment  has  been  slightly  more  prevalent  in  both  Germany  and  the  UK  under  the  economic 
recession  of  the  1980s  than  in  the  1990s,  while  we  observe  similar  levels  of  unemployment  in  France 
for  both  points  in  time  (cf.  also  OECD  1996).  Judged  from  the  duration  of  current  unemployment,  it  is 
not  just  the  level  of  unemployment  which  differs  between  the  countries  but  also  the  seriousness  of 
market  exclusion  faced  in  the  transition  from  education  to  work.  Clearly,  concerns  for  exclusion  are 
appropriate  at  this  early  career  stage  as  the  evidence  on  extended  periods  of  unemployment  shows: 
summing  the  proportions  of  those  long-term  unemployed  and  those  having  returned  to  unemployment 
within  the  last  year,  between  roughly  one  third  of  the  unemployed  in  Germany  (also  assuming  some 
10%  of  recurrent  unemployment)  and  up  to  half  of  the  unemployed  in  the  UK  are  facing  serious 
problems  in  becoming  integrated  into  the  labour  market.  It  is  remarkable  that  even  against  a very 
favourable  overall  situation,  the  picture  is  especially  favourable  with  respect  to  young  Germans 
entering  the  labour  market,  for  whom  unemployment  - if  it  occurs  - tends  to  be  shorter  than  in  both 
France  and  the  United  Kingdom.  In  turn,  the  severity  of  early  exclusion  tends  to  be  highest  in  the  UK. 
Even  with  a slight  convergence  in  these  indicators  over  the  last  decade,  these  stylised  facts  hold  for 
both  the  mid-1980s  and  the  mid-1990s. 

Apart  from  these  differences  in  the  level  and  severity  of  the  youth  unemployment  problem,  the  simple 
breakdown  of  the  situation  before  current  unemployment  reveals  an  interesting  difference  in  the  way 
unemployment  risks  occur  in  the  early  career.  The  distinction  among  the  currently  unemployed 
between  labour  force  entrants  and  experienced  young  unemployed  is  helpful  in  this  respect  as  it  taps 
an  important  distinction  which  will  be  followed  in  more  detail  later.  Specifically,  it  triggers  the  distinction 
between  the  two  major  components  of  unemployment  in  early  labour  market  careers,  namely  search 
for  first  employment  and  instability  of  initial  employment  contracts.  As  is  apparent  from  table  2,  both 
components  are  important  in  the  three  countries:  obviously,  there  is  more  to  unemployment  risk  in 
early  employment  careers  than  just  difficulties  in  locating  a first  job.  In  all  countries,  a significant 
proportion  of  experienced  unemployed  became  unemployed  after  initial  work  involvement  due  to  the 


instability  of  the  initial  employment  found.  Of  course,  the  relative  importance  of  either  type  of 
unemployment  directly  depends  on  the  observation  window  chosen  here,  thus  any  interpretation  at 
face-value  has  to  be  avoided.  Nevertheless,  from  the  estimates  given  in  table  2,  it  seems  that  the 
relative  importance  of  the  search  component  increased  over  the  last  decade  in  the  United  Kingdom, 
while  it  decreased  in  France. 

Table  2:  Unemployment  Patterns  in  the  Transition  from  Education  to  Work 


France 

Germany 

United  Kingdom 

1984 

1994 

1982 

1995 

1984 

1994 

% 

% 

% 

% 

% 

% 

Unemployment  Rate 

22,8 

23,5 

8,6 

6,1 

19,6 

17,9 

Long-Term  Unemployment 

33,7 

29,7 

15,6 

22,5 

46,2 

40,0 

Recurrent  Unemployment 

10,9 

10,4 

N/A 

N/A 

7,8 

8,1 

Reasons  for  current 
unemployment 

* 

Labour  Force  Entrants 

39,3 

32,0 

33,4 

35,5 

41,1 

47,8 

Given  Up  Self-Employment 

0,6 

0,5 

0,7 

0,9 

2,9 

, 3,4 

Former  Employees 

60,1 

67,5 

65,9 

63,6 

56,0 

48,8 

- Thereof:  Had  Temporary 
Contract 

53,8 

65,5 

N/A 

24,7 

20,7 

14,1 

N Unemployed 

2.956 

2.726 

3.068 

1.096 

2.825 

2.011 

Notes  on  definitions  applied: 

Long-term  unemployment  defined  as  the  fraction  of  current  unemployment  lasting  12  months  or  longer; 
recurrent  unemployment  defined  as  the  fraction  among  the  currently  unemployed  whose  unemployment  lasts  less  than  12 
months,  but  who  were  in  unemployment  1 year  ago  as  well;  this  figure  could  not  be  estimated  from  the  German  data  due  to 
lacking  information  on  status  one  year  ago; 

reasons  for  current  unemployment  are  exclusive,  only  the  major  alternatives  are  presented; 

Sources: 

Enquete  Emploi  1984  and  1994;  Mikrozensus  1982  and  1995;  UK  Labour  Force  Survey  1984  and  1994; 

Entrants  into  the  labour  force,  unweighted  results 


Investigating  briefly  the  type  of  employment  held  by  the  experienced  unemployed,  however,  leads  to 
the  recognition  of  a remarkable  difference  in  the  initial  unstable  employment  experiences  of  youth  in 
the  three  countries.  In  all  countries,  the  role  of  self-employment  in  preceding  the  occurrence  of  current 
unemployment  is  negligible.  A significant  difference  does  emerge,  however,  among  former  employees. 
Here  the  major  contrast  is  between  France  on  the  one  hand  and  Britain  and  West  Germany  on  the 
other.  In  comparative  perspective,  the  very  high  percentage  of  54%  (1984)  up  to  66%  (1994)  of  former 
employees  in  France  who  entered  unemployment  because  of  the  termination  of  a prior  temporary 
contract  is  unparalleled  in  the  two  other  economies.  It  appears  that  this  relates  to  the  French  system  of 
alternance  between  participation  in  labour  market  programmes  and  open  unemployment  and  the 
prevalence  of  fixed-term  contracts  at  the  stage  of  entry.  To  sum  up,  this  short  overview  has  provided 
some  indication  that  youth  unemployment  differs  between  the  three  societies  in  many  important 
respects:  it  derives  from  more  than  simple  differences  in  the  levels  of  unemployment,  but  there  are 
also  issues  of  differences  in  the  severity  of  exclusion  as  well  as  with  respect  to  the  relative  importance 


i 


49 


1.13 


of  search  unemployment,  initial  employment  experiences  and  the  instability  of  initial  employment.  We 
will  pursue  some  of  the  implications  in  subsequent  analyses.  As  an  intermediate  step,  we  will  proceed 
to  present  comparative  results  on  educational  differentials  in  unemployment  risks. 


5 Unemployment  Risks  and  Education 

The  core  assumption  underlying  our  study  is  that  the  incidence  of  unemployment  is  highly  related  to 
individuals’  educational  achievement.  In  most  countries,  this  is  a firmly  established  empirical  regularity 
(specifically  for  the  case  of  youth  unemployment  see  e.g.  Franz  et  al.  1997;  Handl  1996;  Helberger  et. 
al.  1994;  Winkelmann  1996;  Ashton/Sung  1992;  Bynner/Roberts  1991;  Evans/Heinz  1994;  Ruiz- 
Quintanilla/Claes  1996).  In  our  study,  the  focus  of  interest  will  therefore  be  on  how  education  affects 
the  incidence  of  unemployment  in  the  initial  labour  market  years  and  on  the  similarities  and  differences 
in  this  respect  between  the  countries.  Furthermore,  we  explore  how  educational  differentials  in 
unemployment  risks  have  evolved  over  the  last  decade.  In  the  presentation  of  our  findings,  we  will 
focus  first  on  current  differences  between  the  countries,  while  the  issue  of  trends  in  differentials  will  be 
taken  up  afterwards. 

As  a starting  point  for  our  analyses,  figure  1 below  provides  evidence  on  the  educational  stratification 
of  unemployment  risks  in  the  three  countries.  The  figure  presents  empirical  results  in  terms  of  both  the 
qualification  specific  unemployment  rates  and  the  relative  advantages  provided  by  different  types  of 
education  with  respect  to  the  incidence  of  unemployment.  Absolute  unemployment  rates  are  shown  in 
the  panels  on  the  left,  while  results  on  the  competitive  advantages  provided  by  qualifications  are 
presented  in  the  panels  on  the  right.  The  latter  are  given  in  terms  of  reciprocal  odds  ratios  of 
unemployment  incidence  among  different  educational  groups  as  compared  to  risks  of  those  with 
compulsory  education  only.  That  is,  the  higher  the  effects  shown,  the  smaller  the  relative  risk  of 
unemployment  among  leavers  with  a certain  educational  background  as  compared  to  the  lowest 
qualified.  These  estimates  are  derived  from  logistic  regression  models  which  only  include  education 
and  a gender  main  effect  (cf.  Appendix  A2).12 

The  analyses  reveal  both  cross-national  similarities  and  dissimilarities.  The  commonalities  between 
the  three  societies  refer  to  five  aspects  of  the  observed  stratification  patterns:  (1)  the  occurrence  of 
very  high  unemployment  rates  for  the  least  qualified  labour  market  entrants,  (2)  substantially  lower 
unemployment  rates  for  tertiary-level  graduates,  (3)  the  relative  advantages  provided  by  tertiary 
education  within  the  different  education  and  training  systems,  (4)  an  inverse,  although  not  necessarily 
linear  relation  between  the  level  of  general  education  and  unemployment  risks,  and  (5)  the  role  of 
general  versus  vocational  qualifications  at  the  different  stages  of  secondary  education. 

The  high  unemployment  rates  faced  by  the  least  qualified  leavers  (CASMIN  lab)  compared  to  fairly 
low  unemployment  rates  among  tertiary  level  leavers  (CASMIN  3a,  3b)  are  apparent  from  figure  1. 


The  figures  present  estimates  in  terms  of  relative  advantages  provided  by  different  levels  and  types  of 
education.  To  provide  a reading  example:  the  figures  give  the  relative  advantage  in  terms  of  unemployment  risks 
as  compared  to  the  lowest  qualified.  That  is,  unemployment  risks  of  French  university  graduates  (CASMIN  3b)  are 
roughly  7 times  less  than  those  of  the  lowest  qualified  (CASMIN  lab).  The  figures  can  be  gained  by  inverting  the 
exponentiated  logit  coefficients  from  Appendix  A2. 


The  qualificational  breakdown  of  youth  unemployment  rates  reveals  - given  the  results  presented  in 
table  2 above  - a rather  unexpected  cross-national  similarity:  across  countries,  there  is  only  small 
variation  in  unemployment  rates  among  both  the  least  qualified  (between  some  33%  in  Germany  and 
45%  in  France)  and  tertiary-level  graduates  (around  5%,  except  some  10%  in  France).  This  also 
implies  similar  relative  returns  to  tertiary  education  in  terms  of  unemployment  risks:  in  all  three 
countries,  higher  education  graduates  incur  seven  to  eight  times  lower  unemployment  risks  as 
compared  to  the  least  qualified  market  entrants.  In  other  words,  the  competitive  advantages  provided 
by  higher  education  are  both  pervasive  and  fairly  similar  in  the  three  European  societies. 

Moreover,  our  results  confirm  a first  hypothesis  outlined  in  the  introduction:  level  of  education  and 
vocational  specificity  of  the  degree  matters.  Firstly,  the  results  clearly  show  an  inverse  relationship 
between  the  level  of  general  education  achieved  and  the  relative  risks  of  unemployment  in  all  three 
societies  (cf.  hypothesis  HI).  Having  obtained  intermediate  secondary  education  (CASMIN  2b)  rather 
than  only  compulsory  education  (CASMIN  lab),  and  obtaining  full  secondary  education  (CASMIN  2c) 
rather  than  intermediate  secondary  education,  and,  finally,  obtaining  a tertiary  degree  (CASMIN  3a, 
3b)  rather  than  secondary  education  only,  respectively  imply  a reduction  of  unemployment  risks.  The 
extent  of  these  relative  advantages  is  also  similar  rather  than  disparate  across  countries.  Secondly,, 
we  also  observe  that  participation  in  vocational  training  at  the  secondary  level  (CASMIN  1c,  2a,  2c 
voc)  pays  off  in  almost  all  cases  in  terms  of  lower  unemployment  risks  as  compared  to  general 
education  at  the  same  level  of  qualification  (CASMIN  lab,  2b,  2c;  cf.  hypothesis  HI). 

Nevertheless,  this  short  discussion  of  similarities  glosses  over  two  crucial  and  related  differences  in 
the  extent  and  pattern  of  stratification.  These  differences  relate  to  both  the  extent  of  competitive 
advantage  provided  by  different  credentials  as  well  as  to  the  more  detailed  pattern  of  advantages, 
especially  in  the  case  of  vocational  training  at  the  secondary  level.  To  treat  the  extent  of  educational 
stratification  first,  we  calculated  a very  rough  indicator  of  the  inequality  of  unemployment  risks 
between  educational  groups.  This  simple  additive  measure  for  deviation  from  a uniform  distribution  of 
unemployment  risks  is  given  in  the  graphs  and  indicates  the  strongest  stratification  of  unemployment 
risks  in  Germany.13  The  British  and  especially  the  French  distributions  appear  much  less  stratified  by 
education,  although  substantial  differentials  are  still  apparent. 


13  This  indicator  is  calculated  from  estimated  odds  ratios  of  unemployment  for  all  educational  groups  relative  to 
CASMIN  lab.  The  total  deviation  is  given  by  I[ori  - 1],  with  ori  = eb  for  b > 0,  and  ori  = (eb)-1  for  b < 0,  across 
educational  groups  CASMIN  1c  to  CASMIN  3b 

O 


1.15 


151 


Figure  1:  Unemployment  Risks  and  Educational  Qualifications 


absolute  unemployment  risks 


France 

relative  protection  against  unemployment 


absolute  unemployment  risks 


United  Kingdom 

relative  protection  against  unemployment 


absolute  unemployment  risks 


West  Germany 

relative  protection  against  unemployment 


| □ 1984  (Germany  1982)  B 1994  (Germany  1995)  j j 1984  {Germany  1982)  ■ ♦ 1 - 1994  (Germany  1995)  | 

Notes: 

Relative  effects  of  education  are  expressed  as  reciprocal  effect  parameter  estimates  compared  to  CASMIN  lab  qualifications; 

see  Appendix  1 for  details  on  logit  models;  dotted  lines  indicate  insignificant  marginal  effect  changes 

Sources: 

Enquete  Emploi  1984  and  1994;  Mikrozensus  1982  and  1995;  UK  Labour  Force  Survey  1984  and  1994; 

Entrants  into  the  labour  force,  unweighted  results 


Moreover,  the  overall  patterns  of  educational  stratification  are  very  different  across  countries  once 
attention  is  paid  to  the  differences  in  the  value  of  secondary  level  vocational  training  (cf.  hypotheses 
H3,  H4).  To  take  the  British  pattern  as  a starting  point,  we  observe  an  almost  linear  relation  between 
qualifications  and  unemployment  risks.  This  pattern  reflects  a straightforward  valuation  of  the  level  of 
education  attained:  both,  each  additional  stage  of  qualification  and  each  vocational  specialization 
within  educational  stages  significantly  reduces  the  unemployment  risks  of  young  people.  The  British 
pattern  contrasts  very  clearly  with  the  pictures  obtained  in  both  other  cases.  In  the  French  case  we 
basically  observe  a three-layered  stratification  by  educational  level : high  unemployment  risks  for  those 
with  only  compulsory  education,  intermediate  relative  risks  for  leavers  from  secondary  education  (with 
some  advantages  to  vocational  qualifications),  and  lowest  risks  for  tertiary  education  graduates.14 
Indeed,  a substantial  decrease  of  unemployment  risks  with  the  achievement  of  academic  education 
seems  to  be  a characteristic  of  the  French  situation.  Still,  there  is  a much  more  striking  contrast 
between  those  latter  countries  and  the  German  pattern  of  stratification:  unemployment  risks  in 
Germany  appear  to  be  polarised  between  those  having  obtained  general  secondary  qualifications  only 
and  those  who  passed  either  vocational  training  or  academic  education.  Unemployment  risks  for  the 
latter  groups  are  fairly  low,  while  the  former  face  substantial  unemployment  risks.  That  is, 
unemployment  risks  are  hardly  stratified  according  to  the  level  of  education  achieved  but  rather  by  the 
distinction  between  vocationally  specific  education  and  training  versus  general  secondary  education 
only.  Essentially,  leavers  from  vocational  training  tracks  at  the  secondary  level  in  Germany  face  similar 
low  or  in  some  cases  even  lower  unemployment  risks  than  university  leavers.  This  finding  is 
unparalleled  by  the  patterns  in  both  Britain  and  France.  Indeed,  it  seems  fair  to  conclude  that  countries 
differ  most  with  respect  to  the  effectiveness  of  vocational  training  (hypotheses  H3,  H4).  This  holds 
both  on  the  lower  CASMIN  1c  (with  the  British  system  providing  the  least  returns)  and  on  CASMIN  2a 
and  2c  voc  level  (with  the  French  leavers  having  the  relatively  least  advantageous  position  across 
countries). 


Changes  over  Time 

Figure  1 also  provides  visual  evidence  of  changes  in  the  educational  stratification  of  unemployment 
risks  over  the  last  decade.  Across  countries,  no  secular  trend  with  regard  to  changes  in  the 
relationship  between  education  and  unemployment  over  time  is  apparent.  Rather,  we  find  significant 
national  variation  in  terms  of  change  over  time.  The  British  pattern  of  stratification  has  remained 
relatively  stable  between  the  mid-1980s  and  the  mid-1990s.  For  France  and  Germany,  opposite  trends 
are  observable:  While  the  educational  stratification  of  youth  unemployment  has  clearly  declined  in 
France,  it  has  become  more  polarised  in  Germany. 

Changes  over  time  are  most  significant  on  the  upper  secondary  and  tertiary  level  of  the  education  and 
training  system.  While  the  relative  returns  to  upper  secondary-  and  tertiary-level  certificates  have 
clearly  declined  in  France,  they  have  increased  in  Germany.  In  France  in  1980,  the  baccalaureat 
implied  substantial  advantages  over  other  secondary  level  qualifications,  general  and  vocational.  Also, 


14  Although  there  is  additional  variation  between  different  qualifications  on  the  CASMIN  1c  - CASMIN  2c  voc  level 
that  conforms  to  the  above  description  of  similarities  between  the  countries,  it  seems  fair  to  draw  this  conclusion 
because  of  the  - in  comparative  perspective  - very  slight  ‘value  added'  of  achieving  baccalaureat  qualifications 
(CASMIN  2c  and  2c  voc)  with  respect  to  unemployment  risks. 


1.17 


the  benefits  pertaining  to  tertiary-level  degrees  were  much  higher  than  in  the  mid-1990s.  In  Germany 
in  1980,  on  the  contrary,  graduating  from  the  tertiary  education  or  completion  of  apprenticeship 
training  following  at  least  a Mittlere  Reife  qualification  (CASMIN  2a/b  level)  held  fewer  competitive 
advantages  over  other  school-leavers  than  ten  years  later.  It  seems  as  if,  beyond  achievement  of  a 
vocational  qualification,  achieving  at  least  the  secondary  intermediate  level  of  education  has  gained  in 
importance  over  time.  Summarising  the  evidence,  we  thus  find  distinctive  trends  in  the  educational 
stratification  of  unemployment,  apart  from  the  cross-country  differences  in  patterns  themselves: 
increased  polarisation  in  the  German  case,  continuity  in  the  United  Kingdom,  and  a substantial 
levelling  out  of  competitive  advantages  pertaining  to  higher  quality  degrees  in  France. 


6 Education  and  Unemployment  in  Labour  Market  Entry  Processes 

As  the  final  step  in  our  analyses,  we  now  explore  the  nature  of  educational  effects  during  labour 
market  entry  processes  in  more  detail.  Therefore,  we  distinguish  between  two  stages  within  the 
transition  period:  first,  the  search  for  initial  employment  after  leaving  education  and  training  and 
second,  the  early  career  stage  after  initial  employment  experience.  We  thus  decompose 
unemployment  risks  in  the  transition  from  education  to  work  into  two  aspects,  namely  access  to  first 
employment  and  instability  of  initial  employment.  The  rationale  favouring  this  setup  is  that  it  will  allow 
an  investigation  of  different  aspects  of  educational  effects  in  the  transition  process:  unemployment 
may  result  from  an  inability  to  access  employment  or  instability  of  initial  employment  itself,  with  the 
decomposition  enabling  us  to  investigate  different  educational  effects  in  either  stage.  The  technical 
setup  of  this  model  has  been  detailed  in  section  3 above;  estimation  results  are  given  in  tabular  form 
in  table  3 below.  Figures  2 and  3 presented  below  provide  graphical  information  on  core  results  in 
terms  of  the  marginal  effects  of  education.  Results  for  this  sequential  model  will  be  discussed  in  the 
following,  naturally  focusing  on  the  effects  of  qualifications. 


First  Entry  into  the  Labour  Market 

The  first  stage  of  our  sequence  model  describes  the  smoothness  of  labour  market  entry:  we  predict 
the  probability  of  ever  having  had  a job  dependent  on  gender,  education  and  time  since  leaving 
education  and  training.  Essentially,  it  is  the  two  latter  effects  that  are  of  key  interest  at  this  stage, 
namely  the  nature  of  educational  advantages  and  the  waiting  time  involved  in' accessing  the  first  job. 
The  educational  effects  of  course  describe  the  nature  of  educational  stratification  in  finding  one’s  first 
job,  while  the  time  effects  allow  for  an  assessment  of  the  immediacy  of  market  entry  in  terms  of 
waiting  time  until  initial  employment.  From  both  the  tabular  and  the  graphical  display  of  respective 
results  (cf.  the  lower  part  of  table  3,  respectively  the  left  and  middle  panel  of  figure  2),  two  country 
patterns  emerge  at  this  stage  of  the  model:  in  terms  of  the  stratification  pattern  of  market  entry,  France 
and  the  United  Kingdom  appear  broadly  similar,  while  the  German  pattern  is  markedly  different  (see 
hypotheses  H3  and  H4). 

In  Germany,  labour  market  entry  is  found  to  occur  fairly  quickly  and  smoothly  after  completion  of  initial 
education  and  training.  Effects  of  time  since  leaving  the  education  and  training  system  on  the 
probability  of  accessing  a first  job  are  hardly  discernible.  That  is,  young  people  just  one  year  out  of 


education  and  training  are  as  likely  to  have  found  their  first  job  as  their  older  counterparts.  This  feature 
comes  with  a strong  educational  stratification  of  labour  market  entry,  however.  Again,  the  educational 
effects  follow  the  by  now  well-known  polarised  pattern:  leavers  from  vocational  training  tracks  on 
secondary  level  (CASMIN  1c,  2a,  2c  voc)  and  tertiary  level  graduates  (CASMIN  3a,  3b)  hardly  face 
any  problems  in  immediate  labour  market  integration.  In  contrast,  leavers  having  obtained  general 
secondary  education  certificates  only  (CASMIN  lab,  2b,  2c),  experience  substantial  entry  problems.  In 
summary,  problems  of  access  to  employment  are  very  much  confined  to  the  least  qualified  in 
Germany,  while  there  is  no  exclusion  of  the  most  recent  entrants. 

Contrasted  with  the  German  pattern,  success  in  finding  a first  job  in  France  and  the  United  Kingdom 
first  of  all  appears  to  be  much  more  related  to  search  time.  In  addition  to  time  effects,  educational 
stratification  is  also  present,  although  the  patterns  are  less  polar  and  suggest  different  rankings  of 
qualifications.  The  strong  effects  of  time  since  leaving  education  and  training  provide  some  evidence 
for  significantly  less  smooth  initial  transitions.  In  their  first  and  second  year  in  the  labour  market,  a 
substantial  proportion  of  the  British  and  French  entrants  are  still  looking  for  their  first  job.  In  contrast  to 
the  German  case,  the  probability  of  finding  a first  job  is  strongly  related  to  and  increasing  with  (fairly 
lengthy)  spans  of  time.  In  that  sense,  the  process  of  labour  market  integration  is  much  more  gradual  in 
these  economies  than  in  Germany  (cf.  hypothesis  H4).  On  the  other  hand,  this  process  is  less 
excessively  structured  and  polarised  by  educational  qualifications:  it  is  self-evident  from  the  French 
and  British  patterns  that  education  does  provide  competitive  advantages  in  accessing  a first  job,  yet 
the  patterns  of  stratification  appear  quite  different.  In  both  cases,  it  is  much  more  the  level  of  education 
that  pays  off  in  terms  of  a smooth  transition  into  the  first  job  than  in  Germany.  Following  an 
educational  career,  almost  each  additional  qualification  entails  a higher  entry  probability  in  Britain  and 
France.  This  contrasts  very  sharply  with  the  German  pattern,  where  leavers  from  apprenticeship 
tracks  fare  slightly  better  than  university  graduates  in  the  transition  from  education  to  work  (cf. 
hypothesis  H3). 

Acquisition  of  Positional  Resources  and  their  Effects 

Once  employment  has  been  located,  unemployment  risks  in  the  subsequent  career  are  subject  to  both 
an  indirect  and  a direct  effect  of  education.  Unemployment  risks  clearly  depend  upon  education  in  the 
sense  that  education  is  continuously  being  used  as  a screening  device  to  discriminate  between 
individuals  in  the  case  of  dismissal  and  recruitment.  On  the  other  hand,  the  risk  of  unemployment  is 
also  related  to  the  type  of  job  position  one  holds.  The  more  stable  and  long-term  one’s  current 
contract,  the  smaller  medium-term  unemployment  risks  are  to  be  expected.  Insofar  as  education  is 
linked  to  issues  of  access  to  stable  positions,  this  reconversion  of  education  into  job  positions  yields 
an  indirect  effect  of  qualifications  on  unemployment  risks.  This  nexus  between  educational 
qualifications  and  job  quality  as  well  as  the  impact  of  different  institutional  contexts  have  been 
extensively  discussed  and  reasonably  well  established  in  comparative  research  (e.g.  Mufler/Shavit 
1998;  Brauns/Muller/Steinmann  1997;  Kerckhoff,  forthcoming  among  others).  And  although  we  do  not 
explicitly  model  this  aspect  in  the  current  model,  we  clearly  observe  its  consequences.  Educational 
effects  in  the  second  stage  of  our  sequence  model  indeed  decline  once  positional  information  in  terms 
of  segment  position  and  type  of  contract  are  included  in  the  equation. 


ERIC 


1.19  # 


Table  3:  Search  and  Instability  in  the  Labour  Market  Entry  Process,  Sequential  Logit  Estimation 


France 

United  Kingdom 

West  Germany 

1984 

1994 

1984 

1994 

1982 

1995 

Unemployment  Risk  after  First  Job 

Women 

0,05  (.06)  "*• 

-0,06  (.06)  "■* 

-0,22  (.06) 

-0,49  (.07) 

0,12  (.07) n 9 

-0,14  (.09) n> 

Educational  Qualifications 
-CASMIN  1c 

-0,60  (.08) 

-0,63  (.09) 

-0,32  (.14) 

-0,20  (.11) 

-0,67  (.09) 

-1,19  (.14) 

- CASMIN  2b 

-0,43  (.10) 

-0,49  (.12) 

-0,63  (.08) 

-0,60  (.10) 

-0,79  (.17) 

-0,88  (.24) 

-CASMIN  2a 

-0,74  (.09) 

-0,72  (.09) 

-0,46  (.11) 

-0,82  (.12) 

-1,12  (.12) 

-1,74  (.15) 

- CASMIN  2c 

-0,81  (.13) 

-0,60  (.13) 

-1,01  (.14) 

-1,15  (.16) 

-1,14  (.22) 

-1,91  (.29) 

- CASMIN  2c  voc 

-1,21  (.14) 

-0,91  (.11) 

-0,96  (.22) 

-1,71  (.24) 

-0,59  (.16) 

-1,84  (.19) 

- CASMIN  3a 

-1,71  (.17) 

-1,19  (.11) 

-0,98  (.23) 

-1,45  (.21) 

-0,82  (.24) 

-1,73  (.26) 

-CASMIN  3b 

-1,91  (.20) 

-1,27  (.13) 

-1,15  (.17) 

-1,51  (.16) 

-0,57  (.21) 

rl , 1 2 (.21) 

Job  Position  Attained 
- Professional  Employment 

-0,63  (.12) 

-0,26  (.10) 

-1 ,15  (.14) 

-1,01  (.13) 

-1,22  (.16) 

-0,86  (.16) 

- Skilled  Employment 

-0,36  (.06) 

-0,02  (.06) 08 

-0,42  (.07) 

-0,54  (.08) 

-0,42  (.08) 

-0,29  (.10) 

- Self-employment 

-1,75  (.27) 

-1,69  (.33) 

-0,67  (.15) 

-0,47  (.15) 

-1,63  (.29) 

-1,20  (.34) 

Temporary  Job 

2,08  (.06) 

1,91  (.06) 

0,91  (.09) 

0,23  (.11) 

N/A 

-0,20  (.11)n# 

Intercept 

-1,46  (.06) 

-1,51  (.07) 

-1,42  (.06) 

-0,93  (.08) 

-2,18  (.09) 

-1,42  (.12) 

N 

12.725 

11.472 

12.642 

10.095 

32.809 

17.355  * 

Log-Likelihood  H0 

-5.325,34 

-5.184,90 

-3.996,03 

-3.158,04 

-5.405,62 

-2.736,90 

Log-Likelihood  H i 

-4.282,28 

-4.294,93 

-3.714,30 

-2.884,41 

-5.224,32 

-2.605,48 

Likelihood  Ratio  Test  / G2 

2.086,12  (12) 

1.779,95(12) 

563,46(12) 

547,26  (12) 

362,60  (11) 

262,84  (12) 

RV 

0,15 

0,14 

0,04 

0,05 

0,01  , 

0,02 

BIC* 

-1.972,70 

-1.667,77 

-450,12 

-436,62 

-248,22 

-145,70 

Attainment  of  First  Job 
Women 

-0,69  (.06) 

-0,23  (.07) 

0,45  (.06) 

0,67  (.08) 

-0,26  (.12) 

-0,33  (.11) 

Educational  Qualifications 
-CASMIN  1c 

1,07  (.08) 

1,21  (.10) 

0,83  (.15) 

0,46  (.11) 

3,60  (.24) 

2,89  (.19) 

-CASMIN  2b 

0,93  (.10) 

1,00  (.12) 

0,99  (.07) 

1,13  (.11) 

0,63  (.20) 

0,82  (.20) 

- CASMIN  2a 

1,56  (.10) 

1,63  (.11) 

1,55  (.14) 

1,61  (.15) 

4,09  (.20) 

3,56  (.19) 

-CASMIN  2c 

1,38  (.12) 

1,44  (.15) 

1,10  (.11) 

1,43  (.14) 

0,72  (.13) 

1,64  (.22) 

-CASMIN  2c voc 

1,88  (.15) 

1,91  (.12) 

1,97  (.27) 

1,52  (.18) 

3,09  (.29) 

3,91  (.31) 

-CASMIN  3a 

2,46  (.16) 

2,50  (.14) 

1,92  (.25) 

1,90  (.21) 

2,04  (.22) 

2,66  (.25) 

-CASMIN  3b 

2,42  (.17) 

2,48  (.14) 

1,81  (.14) 

1,94  (.13) 

2,07  (.16) 

2,45  (.20) 

Time  in  the  labour  market 
- First  year 

-2,22  (.07) 

-3,04  (.05) 

-1,19  (.07) 

-2,66  (.09) 

-1,40  (.15) 

-0,71  (.19) 

- Second  year 

-1,51  (.08) 

-1,70  (.06) 

-0,14  (.09)"8 

-1,67  (.11) 

-0,76  (.24) 

-0,56  (.19) 

- Third  year 

-0,72  (.09) 

-0,73  (.07) 

-0,04  (.10)na 

-1,27  (.12) 

0,81  (.23) 

-0,60  (.15) 

Intercept 

2,28  (.06) 

2,16  (.05) 

1,59  (.05) 

2,21  (.09) 

2,59  (.24) 

1,83  (.11) 

N 

14.414 

12.867 

14.269 

11.178 

34.576 

18.001 

Log-Likelihood  H0 

-5.207,26 

-4.415,86 

-4.706,57 

-3.430,07 

-4.254,44 

-1.784,15 

Log-Likelihood  H i 

-4.256,66 

-3.276,80 

-4.245,62 

-2.728,59 

-3.203,61 

-1.522,05 

Likelihood  Ratio  Test  / G2 

1.901,20  (11) 

2.278,11  (11) 

921,89(11) 

1.402,95  (11) 

2.101,66(11) 

524,20  (11) 

RV 

0,12 

0,16 

0,06 

0,12 

0,06 

0,03 

BIC' 

-1.795,86 

-2.174,03 

-816,67 

-1.300,41 

-1.986,70 

-416,42 

Notes: 

Model  on  Attainment  of  first  job  temporally  precedes  model  on  unemployment  risks  after  first  job;  the  latter  model  includes  only 
the  subsample  of  experienced  individuals; 

Standard  errors  in  parantheses;  n.s.  signifies  statistical  significance  at  p > .05; 

Reference  categories  are:  CASM1N  lab  for  education;  unskilled  employment  for  job  position  attained  and  fourth  year  or  later  in 
case  of  time  in  the  labour  market; 

Sources:  Enquete  Emploi  1984  and  1994;  Mikrozensus  1982  and  1995;  UK  Labour  Force  Survey  1984  and  1994; 
entrants  into  the  labour  force,  unweighted /epults 


156 


1.20 


Figure  2:  Educational  Effects  on  Unemployment,  Sequential  Logit  Model  Estimates 


MID-1990S 

A)  Probability  of  First  Job  B)  Probability  of  First  Job  C)  Unemployment  Risk  after 

by  Time  in  the  Labour  Market  by  Education  First  Job  by  Education 


MID-1980S 

A)  Probability  of  First  Job  B)  Probability  of  First  Job  C)  Unemployment  Risk  after 

by  Time  in  the  Labour  Market  by  Education  First  Job  by  Education 


-France 


-United 

Kingdom 

-West 

Germany 


Notes: 

Figures  show  dummy  variable  discrete  change  effects  calculated  at  the  means  of  all  other  independent  variables  from 

sequential  logit  model  estimates;  see  table  3 for  tabular  display 

Sources: 

Enquete  Emploi  1984  and  1994;  Mikrozensus  1982  and  1995;  UK  Labour  Force  Survey  1984  and  1994; 

Entrants  into  the  labour  force,  unweighted  results  w *. 

ERIC  157 


As  the  focus  of  the  current  paper  is  on  the  direct  effects  of  education,  however,  we  restrict  ourselves 
to  a very  brief  discussion  of  the  effects  of  employment  positions  attained  (cf.  the  upper  half  of  table  3). 
Broadly  speaking,  professional  and  skilled  workers’  positions  as  well  as  self-employment  usually  imply 
lower  unemployment  risks  as  compared  to  unskilled  employment.  The  advantages  provided  by  skilled 
employment  are  generally  lower  than  those  of  professional  positions  or  self-employment.  Individuals 
employed  on  the  basis  of  temporary  contracts  incur  higher  unemployment  risks  than  those  employed 
on  permanent  contracts,  although  this  disadvantage  has  declined  over  the  last  decade.  While  the 
German  and  British  patterns  of  effects  are  fairly  similar  with  respect  to  the  role  of  positional  resources, 
it  is  the  French  pattern  that  clearly  diverges  this  time.  Two  aspects  of  the  French  pattern  seem 
remarkable:  first,  temporary  contract  positions  in  France  hold  great  disadvantage  in  terms  of 
unemployment  risks  as  compared  to  both  Britain  and  Germany,  while  the  countries  appear  broadly 
similar  as  far  as  unemployment  risks  for  individuals  in  permanent  positions  are  concerned.  Second, 
the  competitive  advantages  provided  by  professional  and  skilled  employee  positions  are  lowest  among 
the  three  countries,  and  this  appears  as  a consequence  of  changes  over  the  last  decade. 


Screening  Effects  after  the  Attainment  of  the  First  Job 

However,  even  after  an  initial  conversion  of  educational  qualifications  into  job  positions  has  occurred, 
we  find  evidence  of  significant  effects  of  education  on  unemployment  risks  in  the  early  labour  market 
career  (cf.  the  upper  half  of  table  3;  hypotheses  H2,  H5).  As  has  been  detailed  above,  one  would  also 
expect  continued  effects  of  education  since  education  may  be  said  to  maintain  its  usefulness  in  the 
evaluation  of  workers  over  an  employment  career.  Essentially,  the  estimates  of  educational  effects  in 
the  second  stage  of  the  labour  market  career  reconfirm  the  observations  made  so  far.  Here,  again  we 
find  evidence  of  two  major  distinct  patterns  of  educational  stratification  of  unemployment  risks  and 
additional  slight  differences  between  France  and  the  UK  (cf.  the  panels  on  the  right  hand  side  of 
figures  2 and  3;  hypothesis  H5).  For  France  and  the  United  Kingdom,  the  patterns  closely  resemble 
the  patterns  already  identified  in  section  5 above:  an  inverse,  almost  linear  relation  between  level  of 
qualification  and  unemployment,  with  the  French  pattern  exhibiting  more  of  a three-layered 
differentiation  between  individuals  having  obtained  compulsory  education  only,  those  having 
completed  secondary  qualifications  and,  particularly  advantaged,  tertiary-level  graduates.  The 
interesting  case  is  Germany  again,  where  educational  differentials  apart  from  the  contrast  of  CASMIN 
lab  versus  all  other  qualifications  are  simply  non-existent  in  terms  of  job  instability  - once  initial 
employment  has  been  found  (cf.  hypothesis  H5).  This  finding  is,  of  course,  in  stark  contrast  to  our 
earlier  results  of  strong  stratification  for  the  transition  period  in  general  and  especially  with  respect  to 
initial  search  in  the  market.  Some  qualifying  remarks  are  necessary,  however,  both  for  details  of  the 
relative  positions  of  single  qualifications  and  on  the  interaction  between  the  two  stages  of  our  model. 

One  appealing  feature  of  the  model  estimated  here  is  the  possibility  of  assessing  reinforcing  versus 
counteracting  effects  of  qualifications  over  the  different  stages  of  the  transition  process.  That  is,  we 
are  able  to  establish  the  extent  to  which  those  qualifications  providing  for  smooth  first-time  entry  also 
imply  relatively  stable  positions  afterwards  or  whether  certain  trade-offs  exist  here.  From  a comparison 
of  educational  effects  at  either  stage  of  the  model  (cf.  the  middle  and  right  panels  of  figures  3 and  4), 
the  general  conclusion  is  that  (dis)advantages  pertaining  to  qualifications  tend  to  be  reinforced  in  the 
course  of  the  labour  market  entry  process.  In  almost  all  cases,  those  qualifications  that  provide 


erJc 


(idt  158 122 


smooth  entry  also  provide  lower  unemployment  risks  afterwards.  Broadly  speaking,  there  is  only  one 
exception  to  this  rule,  namely  the  case  of  intermediary  secondary-level  qualifications  (CASMIN  2a/2b) 
in  Britain.  This  is  the  only  major  case  where  slightly  different  patterns  are  apparent:  in  the  second 
stage  of  the  model,  where  leavers  with  vocational  qualifications  from  the  CASMIN  2a  level  clearly  face 
a disadvantaged  position  as  compared  to  the  general  O-level  type  (CASMIN  2b)  qualifications.  In  fact, 
this  is  the  only  case  where  general  education  qualifications  go  together  with  higher  employment 
stability  as  compared  to  their  vocational  counterparts  on  the  same  level  of  education. 

Breaking  down  unemployment  risks  into  a job  search  and  a job  instability  component,  moreover, 
sheds  some  light  on  the  actual  processes  underlying  the  observed  overall  educational  stratification  of 
youth  unemployment  as  described  in  section  5 above.  Trying  to  summarise  our  comparative  findings, 
it  is  probably  fair  to  conclude  that  two  basic  patterns  of  educational  effects  are  apparent:  the  contrast 
between  Germany  on  the  one  hand,  and  France  and  the  UK  on  the  other.  Taking  the  latter  countries 
first,  substantial  educational  differentials  are  apparent  for  both  stages  of  our  model  - in  addition  to, 
and  above,  controlling  for  positional  resources  and  other  factors.  That  is,  the  overall  educational 
stratification  is  a product  of  reinforcing  (dis)advantages  in  terms  of  both  access  to  employment  and  the 
stability  of  employment  in  the  early  career.  And  although  the  type  of  employment  is  related  tq 
subsequent  unemployment  risk,  educational  effects  continue  to  operate  and  qualifications  retain  their 
comparative  advantages.  Of  course,  there  are  additional  differences  between  France  and  Britain, 
which  we  would  argue  are  deviations  from  a common  pattern.  The  most  striking  of  these  is  the  role 
played  by  temporary  contracts  (partially  including  work  experience  contracts)  in  unemployment  risks. 
While  unemployment  risks  for  individuals  in  temporary  contract  positions  are  slightly  higher  in  the  UK, 
French  youth  in  this  type  of  employment  face  dramatically  higher  risks  of  unemployment.  Certainly,  the 
French  youth  labour  market  is  characterised  by  substantial  volatility  and  instability  partially  linked  to 
the  excessive  use  of  temporary  work  experience  schemes.'5 

Compared  to  these  two  countries,  the  situation  is  fundamentally  different  in  West  Germany.  Here,  the 
very  strong  overall  qualificational  stratification  of  unemployment  is  only  partially  reproduced  by  the 
sequential  model.  Strong  educational  differentials  in  line  with  the  overall  pattern  are  apparent  for  the 
search  stage,  yet  the  educational  stratification  of  employment  instability  is  very  weak.  In  the  second 
stage  of  our  model,  we  only  found  a small  contrast  between  the  least  qualified  and  all  other  leaver 
groups,  with  no  sign  of  further  differentiation.  How  do  these  findings  relate  to  the  overall  stratification 
pattern  observed?  One  might  argue  that  the  very  strong  stratification  observed  results  from  a very 
strong  qualificational  stratification  of  attainment  of  first  job  - which,  moreover,  is  clearly  linked  only  to 
qualifications  without  the  effect  of  time  since  entry  (cf.  hypothesis  H3).  Given  the  substantial 
importance  of  experienced  unemployed  also  in  the  German  sample,  this  explanation  nevertheless 
seems  only  partially  convincing.  It  appears  insufficient  to  account  for  the  lack  of  educational 
stratification  of  exclusion  after  attaining  first  job.  One  possible  explanation,  consistent  with  earlier 
studies  and  aspects  of  our  findings,  centers  on  the  close  match  between  qualifications  and  initial 
employment  in  the  German  labour  market.  If  instability  is  a function  of  the  type  of  employment  and 


15  Certainly,  we  cannot  claim  a causal  treatment  effect  of  fixed-term  contracts  for  France  from  our  analyses:  we 
can,  however,  state  that  fixed-term  contracts  in  France  are  subject  to  extraordinarily  high  risks  of  subsequent 
unemployment  as  compared  to  small  effects  in  both  the  UK  and  West  Germany  - irrespective  of  this  being  related 
to  a causal  effect  or  the  result  of  sharp  selectivity. 


1.23 


159 


qualifications  allow  for  a clear  matching  to  employment  positions,  then  a remaining  single  influence  of 
job  characteristics  may  be  expected  in  the  regressions.  As  convincing  as  this  may  sound,  empirical 
support  is  only  partially  indicative  of  this.  Clearly,  educational  effects  are  very  low  and  even  level  out 
further  once  positional  controls  are  introduced.  However,  as  judged  from  the  goodness-of-fit  indices 
provided,  the  positional  controls  introduced  do  not  appear  to  capture  much  variation  in  unemployment 
risks  themselves.  Of  course,  this  can  be  related  both  to  the  invalidity  of  the  tenet  suggested  or  the 
crudeness  of  the  indicators  adopted.  Further  research  on  this  point  is  obviously  needed. 

Changes  in  Educational  Effects  over  Time 

In  comparing  our  results  for  the  mid-1990s  to  those  established  for  the  mid-1980s,  we  are  additionally 
able  to  establish  evidence  for  changes  over  time  in  educational  effects  on  unemployment  processes. 
With  respect  to  the  first  step  of  finding  first  employment,  we  observe  hardly  any  changes  in  the 
educational  effects  for  any  of  the  three  countries.  Interestingly,  the  time  effects  change  in  the  French 
and  British  cases,  but  not  in  the  German  case:  specifically,  there  are  marked  increases  in  the  negative 
effects  of  the  very  first  year  in  the  market.  Thus,  on  average,  it  seemingly  became  more  difficult  to 
immediately  locate  initial  employment  in  these  two  countries  in  the  1990s.  Somewhat  stronger 
changes  over  time  are,  however,  apparent  for  the  second  stage  of  our  sequential  model.  It  is  evident 
that  the  strong  country  differences  in  the  patterns  of  educational  stratification  observed  for  the  1990s 
are  partly  due  to  divergent  patterns  of  changes  over  the  last  decade.  Changes  in  the  educational 
distribution  of  unemployment  risks  in  the  early  career  took  place  in  all  three  countries.  Yet,  the  pattern 
of  change  differs  between  the  countries:  in  France,  for  once,  we  establish  an  increase  in 
unemployment  risks  biased  towards  higher  level  qualifications.  The  contrasting  case  is  evident  for 
Germany,  where  unemployment  risks  only  increased  substantially  for  the  least  qualified  with  lower  and 
intermediate  general  secondary  qualifications.  For  labour  market  entrants  holding  occupational 
qualifications,  the  situation  remained  basically  unchanged,  leavers  from  CASMIN  2c  voc  and  3a  even 
face  lower  unemployment  risks  in  the  1990s.  The  United  Kingdom,  finally,  probably  shows  the  most 
interesting  pattern  in  this  respect.  As  in  Germany,  unemployment  risks  also  declined  for  higher 
education  leavers,  while  they  increased  for  the  lower  level  qualifications.  However,  it  was  not  only  the 
least  qualified  with  compulsory  education  who  faced  increasing  unemployment  risks,  but  rather  those 
obtaining  slightly  more  demanding  qualifications  on  CASMIN  1c  and  2b  level. 


ERIC 


1.24 


Figure  3:  Trends  in  the  Educational  Stratification  of  Unemployment,  Sequential  Logit  Model 
Estimates 


France 

A)  Probability  of  First  Job  by  Education  B)  Unemployment  Risk  after  First  Job  by  Education 


- -4-  1984 
4 1994 


A)  Probability  of  First  Job  by  Education 


United  Kingdom 

B)  Unemployment  Risk  after  First  Job  by  Education 


West  Germany 

A)  Probability  of  First  Job  by  Education  B)  Unemployment  Risk  after  First  Job  by  Education 


- -4-  1984 
4—1994 


- -4-  1982 
4—1995 


Notes: 

Figures  show  dummy  variable  discrete  change  effects  calculated  at  the  means  of  all  other  independent  variables  from 

sequential  logit  model  estimates:  see  table  3 for  tabular  display 

Sources: 

Enquete  Emploi  1984  and  1994;  Mikrozensus  1982  and  1995;  UK  Labour  Force  Survey  1984  and  1994; 

Entrants  into  the  labour  force,  unweighted  results 


1.25 


161 


7 Summary 

In  this  paper,  we  have  examined  the  on  education  and  unemployment  risks  of  school-leavers’  early 
labour  market  careers.  The  main  objective  was  to  explore  how  young  people's  risk  of  unemployment  is 
related  to  educational  achievement,  and  the  way  in  which  this  relationship  is  shaped  by  the 
institutional  embeddedness  of  the  educational  and  employment  system.  For  this  reason,  a 
comparative  perspective  was  applied  to  Germany,  France  and  the  United  Kingdom,  three  countries 
that  differ  considerably  in  the  characteristics  of  their  national  educational  systems  and  in  the 
organisation  of  their  labour  markets.  Moreover,  an  historical  perspective  on  each  of  the  three  countries 
was  adopted  to  get  some  idea  of  trends  in  the  educational  stratification  of  unemployment  risks  over 
the  past  decade. 

A first  global  glance  at  youth  unemployment  in  the  three  countries  reveals  some  substantial 
differences  in  overall  levels  of  unemployment,  reasons  for  exposure  to  unemployment  as  well  as  in  the 
pervasiveness  of  long-term  unemployment,  search-unemployment  upon  leaving  school  and  the 
instability  of  early  careers.  These  cross-national  differences  tend  to  be  fairly  stable  from  the  mid-1980s> 
to  the  mid-1990s.  Beyond  these  global  differences,  the  three  countries  share  a basic  similarity:  a 
distinctive  educational  stratification  of  unemployment  risks. 

Our  analyses  show  that  young  people's  risk  of  unemployment  is  strongly  related  to  their  educational 
(non-)achievement  in  all  three  countries.  Unemployment  rates  are  typically  highest  among  school- 
leavers  with  compulsory  education  only  and  lowest  among  graduates  from  higher  education.  Despite 
substantial  cross-country  differences  in  national  unemployment  rates,  the  absolute  rates  faced  by  the 
lowest  and  the  highest  qualified  school-leavers,  are  fairly  similar.  This  implies  that  in  all  three 
countries,  tertiary  education  provides  significant  advantages,  and  compulsory  education  major 
disadvantages  with  respect  to  labour  market  integration.  Also,  we  observe  benefits  pertaining  to 
vocational  qualification:  in  all  countries,  vocational  qualifications  significantly  reduce  the  likelihood  of 
unemployment  as  compared  to  having  only  general  education  at  the  same  level.  Overall,  our  findings 
support  our  hypothesis  HI  outlined  in  section  2:  employers  tend  to  reward  two  facets  of  educational 
achievement:  the  hierarchical  level  of  education  reached  and  the  vocational  specificity  of  one's 
education. 

We  have  also  argued  that  employers  use  more  than  just  educational  screening  to  select  among  young 
people,  and  that  the  extent  to  which  educational  resources  make  a difference  should  depend  upon 
young  peoples’  career  stage.  More  precisely,  it  was  suggested  that  the  differentiating  impact  of  formal 
education  should  be  particularly  strong  when  young  people  are  about  to  enter  the  labour  market  (“get 
in  the  door”)  after  leaving  school.  This  effect  should  be  smaller  as  soon  as  they  have  succeeded  in 
finding  employment.  When  entering  the  labour  market,  school  leavers  should  to  some  degree  be  able 
to  transform  their  educational  resources  into  positional  resources  (occupational  position,  type  of 
contract  etc.)  which  more  or  less  protect  them  against  the  risk  of  becoming  unemployed.  At  the  same 
time,  their  performance  on  the  job  provides  employers  with  reliable  indicators  of  their  productivity 
which  can  be  used  to  screen  employees  should  staff  cuts  be  necessary,  rather  than  drawing  on 

162  126 


education  as  an  indirect  measure  of  their  ability.  The  empirical  analysis  in  section  6 supports  our 
hypothesis  H2:  the  educational  stratification  of  unemployment  risks  is  indeed  particularly  strong  at 
labour  market  entry,  that  is  with  respect  to  search-unemployment  immediately  after  leaving  school. 
Still,  educational  achievement  continues  to  matter  after  initial  hiring  into  the  labour  force  has  taken 
place,  particularly  in  terms  of  one’s  chances  of  a stable  career.  As  we  argue  in  section  2,  that  this  may 
occur  because  reliable  information  about  individuals’  productive  capabilities  emerges  slowly  with  time 
in  the  job,  so  that  employers  still  see  formal  education  as  an  important  screening  device  when 
decisions  about  staff  cuts  must  be  made.  Also,  employers  tend  to  reward  credentials  as  a matter  of 
firm  policy  and  for  their  social  value. 

Thurow’s  labour  queue  model  also  suggests  that  the  precise  shape  of  the  labour  queue  depends  upon 
employers’  preferences.  Employers,  in  turn,  are  embedded  in  a specific  institutional  environment  of 
the  organisation  of  labour  markets  and  human  resources  endowed  with  certain  qualifications  and  skills 
that  are  made  available  by  the  national  educational  system.  This  environment  considerably  shapes 
their  preferences  for  certain  credentials  over  others.  Besides  the  commonalties  in  the  educational 
stratification  of  unemployment  risks  that  have  been  outlined,  our  analyses  confirm  considerable  cross- 
national dissimilarity,  in  particular  between  Germany  on  the  one  side,  and  France  and  the  UK  on  the 
other  side,  that  seem  to  be  linked  to  particularities  of  national  institutions.  Basically,  the  countries  differ 
in  the  overall  extent  to  which  education  makes  a difference,  and  - closely  related  to  that  - in  the 
precise  way  in  which  unemployment  risks  are  stratified  by  education,  that  is  in  the  degree  to  which 
each  of  the  two  facets  of  educational  achievement,  level  of  general  education  or  vocational  specificity, 
are  valued.  Relying  on  a rough  indicator  of  the  extent  of  inequality  among  the  various  educational 
groups,  we  find  that  the  strongest  stratification  of  unemployment  risks  according  to  education  prevails 
in  Germany.  This  finding  can  be  put  down  to  the  sharp  skill-divide  between  vocationally  qualified  and 
vocationally  unqualified  school-leavers:  this  confirms  our  hypothesis  H3:  in  the  German  occupational 
labour  market,  vocational  qualifications  provide  a clear  advantage  over  having  only  general  education 
at  every  level.  Vocationally  qualified  school-leavers  profit  from  a smooth,  that  is  transition  into 
employment  immediately  after  completing  their  training.  Moreover,  they  seem  to  benefit  from  job 
allocations  which  provide  them  with  quite  substantial  security  of  employment  during  their  early  career. 
The  opposite  side  to  the  employment  security  offered  to  vocationally  qualified  school-leavers  in 
occupational  labour  markets,  is  the  way  in  which  unqualified  school-leavers  are  excluded  from  the 
labour  market. 

As  expected  in  hypothesis  H4,  in  France  and  the  UK  exposure  to  unemployment  is  less  rigidly 
stratified  according  to  educational  achievement.  The  role  of  education  in  these  countries  is  different  to 
the  situation  in  Germany.  Due  to  the  prevalence  of  firm-internal  rather  than  occupational  labour 
markets,  there  is  no  such  polarisation  of  unemployment  chances  according  to  the  achievement  of 
vocational  qualifications  as  in  Germany.  Yet,  compared  to  Germany,  the  level  of  education  reached  in 
the  school  system  turns  out  to  be  a relatively  more  important  signal  for  securing  employment. 


1.27 


This  occurs  along  with  a generally  less  smooth  transition  into  the  employment  system  in  the  United 
Kingdom  and  France  and,  in  terms  of  young  people’s  chances  of  continued  employment  following 
entry  into  the  labour  market,  formal  education  is  quite  a differentiating  resource.  Hence,  our  analysis 
also  confirms  hypothesis  H5:  we  find  much  weaker  direct  effects  of  formal  education  on  young 
people’s  chances  of  persisting  in  the  labour  market,  once  an  entry  has  taken  place,  in  Germany  than 
in  the  UK  and  France.  In  Germany,  tight  selection  at  entry  into  the  occupationally  structured  labour 
market  implies  a smooth  and,  in  terms  of  job  allocation,  “structured"  transition  into  the  employment 
system  for  those  who  are  endowed  with  the  critical  entry  tickets.  Once  these  school  leavers  enter  the 
closed  system,  they  are  able  to  convert  their  educational  resources  into  a (more  or  less)  beneficial 
labour  market  position  whose  attributes  largely  determine  chances  of  continued  employment.  By 
contrast,  less  rigid  hiring  practices  in  France  and  the  UK  imply  that  transition  into  employment  upon 
completion  of  schooling  is  typically  less  “structured”  in  terms  of  an  institutionalised  “correspondence” 
of  educational  and  occupational  entitlement  at  labour  market  entry  (see  Shavit/Muller  1998).  The 
conversion  of  educational  resources  into  an  adequate  and  safe  job  takes  place  over  a longer  period 
than  in  occupational  labour  markets.  Thus,  one  main  difference  between  Germany  and  the  two  other 
countries  with  regard  to  the  educational  stratification  of  unemployment  risks  is  that  in  the  German 
context  of  occupational  labour  markets,  exclusion  from  entry  into  employment  operates  on  a clear-cut 
qualificational  base.  However,  “underachievement”  in  the  education  system  will  also  be  penalised  in 
the  UK  and  France.  This  penalty  takes  the  form  of  allocation  to  a “bad”  job  and  a temporary 
employment  contract  at  initial  hiring  and,  beyond  that,  a generally  disadvantageous  position  because 
of  their  educational  performance  when  it  comes  to  career  advancement  and  lay-off  decisions. 
Therefore,  the  consequences  of  educational  achievement  in  terms  of  low  unemployment  risks  will 
unfold  later  in  the  course  of  young  people’s  early  careers  than  in  Germany,  with  respect  to  their 
chances  of  career  advancement  and  continued  employment.  Yet,  differences  between  the  two 
countries  exist  in  the  general  importance  of  educational  credentials:  the  historical  tradition  of 
credentialism  in  France  gives  formal  education  an  even  stronger  importance  in  employers’  personnel 
decisions  than  in  the  UK.  However,  the  analyses  suggest  a fairly  clear-cut  notion  of  what  “bad” 
educational  performance  is  in  each  country:  for  the  (dis)advantage  attached  to  certain  achievements 
in  the  educational  system  tend  to  be  reinforced  over  the  course  of  people’s  careers:  those  school- 
leavers  who  have  the  best  chances  of  entering  the  labour  market  also  tend  to  have  the  best  chances 
of  profiting  from  a fairly  stable  early  career  in  terms  of  lower  risks  of  losing  their  job. 

What  is  the  impact  of  educational  expansion  on  the  educational  stratification  of  unemployment  risks? 
Is  there  a common  trend  observable  across  the  three  countries?  The  analyses  reveal  no  spectacular 
changes  in  education  effects  over  the  past  decade.  The  changes  seem  to  reflect  the  idiosyncrasies  of 
the  particular  setting  in  each  country  rather  than  a secular  trend  across  nations.  What  do  the  results 
imply  in  terms  of  the  Thurow  model  put  forward  in  section  2?  At  first  view,  there  does  not  seem  to  be 
much  support  for  the  hypotheses  that  we  have  derived  from  it,  apart  from  - in  some  respects  - the 
position  of  Germany.  In  Germany,  we  observe  that  the  increase  in  qualified  school-leavers  in  the 
course  of  educational  expansion  fosters  indeed  a greater  risk  of  labour  market  exclusion  for  the 
unqualified.  In  other  words,  in  occupational  labour  markets,  educational  expansion  leads  to  an  even 


stronger  closure  against  those  already  being  largely  excluded.  In  the  French  and  British  labour 
markets,  by  contrast,  we  do  not  find  the  same  trend.  In  view  of  the  institutional  reforms  undertaken  in 
the  educational  and  employment  systems,  the  British  and  French  findings  do  not  seem  easily 
comprehensible.  While  the  British,  for  example,  have  made  great  efforts  to  reform  their  vocational 
education  and  training  system  and  to  deregulate  the  labour  market,  we  find  almost  no  indication  of 
improved  relative  chances  for  those  with  vocational  qualifications,  rather  the  opposite  as  far  as  labour 
market  entry  is  concerned.  In  France,  we  indeed  observe  a slight  improvement  in  the  relative  position 
of  vocationally  qualified  school-leavers  over  those  with  general  qualifications  only,  in  particular  with 
regard  to  the  prospects  of  a stable  career,  which  may  come  from  modernisation  of  the  vocational 
education  system.  The  parallel  decline  in  the  relative  benefits  to  tertiary  education,  however,  seem 
counter-intuitive  to  the  hypotheses  that  we  derived  from  the  Thurow  model.  Yet,  more  detailed 
research  on  this  issue  may  provide  further  insight  into  the  validity  of  the  labour-queue  model.  In  the 
case  of  the  educational  stratification  of  unemployment  in  Germany,  we  have  made  use  of  the  idea  of  a 
labour  queue  to  explain  “who  gets  a job,  who  does  not?”,  while  assuming  that  the  jobs  to  be 
distributed  are  more  or  less  homogeneous  in  terms  of  the  unemployment  risk  attached  to  them  after 
people  are  hired  into  the  job.  In  France,  the  assumption  of  homogeneous  clusters  of  jobs  cannot  be 
made.  Due  to  the  predominance  of  internal  labour  markets  and  the  need  to  screen  new  employees  for 
some  period  of  time,  a large  number  of  jobs  are  only  temporary.  These  precarious  job  situations  have 
increased  in  number  over  time  and,  as  the  empirical  results  in  table  2 have  shown,  are  associated  with 
a very  high  risk  of  subsequent  unemployment.  French  employers  have  shown  a strong  tendency  over 
the  past  few  years  to  cope  with  the  uncertain  economic  situation  by  more  flexible  employment 
relationships  which  allows  them  to  respond  more  quickly  and  easily  to  changing  market  demand.  In 
consequence,  if  the  ideas  of  the  labour  queue  model  are  applied  to  explaining  the  educational 
stratification  of  unemployment  on  the  French  labour  market,  then  it  would  need  to  answer  two 
questions:  first,  “who  gets  a job,  who  does  not?”,  second,  “who  gets  a good  job,  who  gets  a bad  job?”, 
the  latter  being  attached  to  a high  risk  of  unemployment  afterwards.  With  respect  to  the  second 
question,  the  idea  of  a labour  queue  would  imply  that  the  lowest  qualified  school-leavers  (who  have 
been  allocated  a job  upon  completion  of  schooling)  are  placed  at  the  top  of  the  queue,  the  higher  at 
the  bottom.  In  times  of  educational  expansion,  the  proportion  of  low  qualified  school-leavers  among  a 
school-leaver  cohort,  and  among  those  who  have  been  offered  a job  in  particular  (see  section  2), 
decreases  substantially.  If  the  number  of  “bad  jobs"  associated  with  a high  risk  of  subsequent 
unemployment  remains  stable  or  even  increases  over  time  as  in  France  (Balsan  et  al,  1998),  then 
more  qualified  school-leavers  should  be  allocated  to  these  bad  jobs  than  before.  Empirical  evidence 
from  other  research  confirms  this.  In  consequence,  school-leavers  with  higher  education  should 
experience  higher  risks  of  unemployment  and,  on  average,  a less  favourable  relative  position 
compared  to  the  lowest  qualified  than  in  earlier  times.  More  extensive  research  is  needed  to  elaborate 
on  this  issue.  Investigating  the  consequences  of  educational  expansion  in  more  detail  requires 
exploring,  for  example,  whether  reform  of  national  education  and  training  systems  has  been 
counterbalanced  by  developments  in  the  youth  labour  market,  and  most  importantly,  empirically 
disentangling  the  impact  of  supply  and  demand-side  factors. 


165 


1.29 


References 

Arrow,  Kenneth  J.,  1973:  Higher  Education  as  a Filter.  Journal  of  Public  Economics  2,  193-216. 


Ashton,  David  N.  and  Jonny  Sung,  1992:  The  Determinants  of  Labour  Market  Transitions:  An 
Exploration  of  Contrasting  Approaches.  Work,  Employment  & Society  6,  1-21. 

Balsan  ,D.,  S.  Hanchane  and  P.  Werquin,  1998:  Sub-minimum  Wage  Employment  and  Earnings 
Profiles  in  Low  Skill  Labour  Market.  Evidence  from  French  Panel  Data;  1989  - 95.  In:  Bazen 
S.,  M.  Gregory  and  W.  Salverda  (eds.)  Low  Wage  Employment. 

Bills,  David  B.,  1988:  Educational  Credentials  and  Promotions:  Does  Schooling  Do  more  than  Get  You 
in  the  Door?  Sociology  of  Education  61,  52-60. 

Blossfeld,  Hans-Peter,  1989:  Kohortendifferenzierung  und  KarriereprozeB.  Eine  Langsschnittstudie 
iiber  die  Veranderung  der  Bildungs-  und  Berufschancen  im  Lebensverlauf.  Frankfurt/Main: 
Campus. 

Bowles  and  Gintis,  1976:  Schooling  in  Capitalist  America:  Educational  Reform  and  the  Contradictions 
of  Economic  Life.  London:  Routledge  and  Kegan  Paul. 

Braun,  Michael  and  Walter  Muller,  1997:  Measurement  of  Education  in  Comparative  Research. 
Comparative  Social  Research  16,  163-201. 

Brauns,  Hildegard,  1998:  Bildung  in  Frankreich.  Eine  Studie  zum  Wandel  herkunfts-  und 
geschlechtsspezifischen  Bildungsverhaltens.  Opladen:  Leske+Budrich. 

Brauns,  Hildegard,  1999  (forthcoming):  Vocational  Education  in  Germany  and  France.  In:  Hans-Peter 
Blossfeld  and  Reinhard  Stockmann  (eds.):  Globalization  and  Changes  in  Vocational  Training 
Systems  in  Developing  and  Advanced  Industrialized  Countries.  Special  Issue  of  the 
International  Journal  of  Sociology. 

Brauns,  Hildegard,  Dietmar  Haun  and  Susanne  Steinmann,  1997.  Die  Konstruktion  eines  international 
vergleichbaren  Klassenschemas  (EGP).  Erwerbsstatistische  Besonderheiten  am  Beispiel  von 
Labour  Force  Surveys  der  Bundesrepublik  Deutschland,  Frankreichs,  Grodbritanniens  und 
Ungarns.  MZES  Working  Paper  AB  1/22.  Mannheim:  Mannheimer  Zentrum  fur  Europaische 
Sozialforschung. 

Brauns,  Hildegard,  and  Susanne  Steinmann,  1997:  Educational  Reform  in  France,  West  Germany,  the 
United  Kingdom  and  Hungary.  Updating  the  CASMIN  Educational  Classification.  MZES 
Working  Paper  AB  1/21 . Mannheim:  Mannheim  Centre  for  European  Social  Research. 

Brauns,  Hildegard  and  Susanne  Steinmann,  1999:  Educational  Reform  in  France,  West-Germany,  the 
United  Kingdom  and  Hungary.  Updating  the  CASMIN  Educational  Classification.  ZUMA- 
Nachrichten,  1999:  im  Druck. 

Brauns,  Hildegard,  Walter  Muller  and  Susanne  Steinmann,  1997:  Educational  Expansion  and  Returns 
to  Education.  A Comparative  Study  for  Germany,  France,  the  UK,  and  Hungary.  MZES 
Working  Paper  AB  I / 23.  Mannheim:  Mannheim  Centre  for  European  Social  Research 

Breen,  Richard,  Damian  F.  Hannan  and  Richard  O’Leary,  1995:  Returns  to  Education:  Taking  Account 
of  the  Employers’  Perception  and  Use  of  Educational  Credits.  European  Sociological  Review 
11,59-73. 

BOchtemann,  Christoph,  JOrgen  Schupp  and  Dana  Soloff,  1994:  From  School  to  Work:  Patterns  in 
Germany  and  the  United  States.  In:  Johannes  Schwarze,  Friedrich  Buttler,  and  Gert  G. 
Wagner  (eds.):  Labour  Market  Dynamics  in  Present  Day  Germany.  Frankfurt,  Boulder: 
Campus,  Westview,  112-141. 


ERIC 


1 3n 


Bynner,  John  and  Ken  Roberts  (eds.),  1991:  Youth  and  Work:  Transition  to  Employment  in  England 
and  Germany.  London:  Anglo-German  Foundation. 

CATEWE  Research  Network,  1998:  A Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe  (CATEWE):  A Conceptual  Framework.  Dublin,  Edinburgh,  Amsterdam,  Maastricht, 
Marseille  and  Mannheim:  Report  to  the  European  Commission. 

Collins,  Randall,  1979:  The  Credential  Society.  A Historical  Sociology  of  Education  and  Stratification. 
New  York:  Academic  Press. 

Drexel,  Ingrid  (ed.),  1993:  Das  Ende  des  Facharbeitertums?:  Neue  mittlere  Bildungs-  und 
Karrierewege  in  Deutschland  und  Frankreich  - ein  Vergleich.  Frankfurt/Main:  Campus. 

Erikson,  Robert  and  John  H.  Goldthorpe,  1992:  The  Constant  Flux.  A Study  of  Class  Mobility  in 
Industrial  Societies.  Oxford,  Clarendon  Press. 

Evans,  Karen,  and  Walter  R.  Heinz  (eds.),  1994:  Becoming  adults  in  England  and  Germany.  London: 
Anglo-German  Foundation. 

Eyraud,  Francois,  David  Marsden  and  Jean-Jacques  Silvestre,  1990:  Occupational  and  Internal 
Labour  Markets  in  Britain  and  France.  International  Labour  Review  129,  501-517. 

Frenzel,  Hansjorg  and  Paul  Luttinger  (1990??):  Die  Mikrozensen  der  70er  und  80er  Jahre  in  der 
Abteilung  Mikrodaten.  ZUMA  - Technischer  Bericht  Nr.  90/08.  Mannzeim:  Zentrum  fur 
Umfragen  Methoden  und  Analysen. 

Franz,  Wolfgang,  Joachim  Inkmann,  Winfried  Pohlmeier,  and  Volker  Zimmermann,  1997:  Young  and 
Out  in  Germany:  On  the  Youth’s  Chances  of  Labour  Market  Entrance  in  Germany.  NBER 
working  paper  6212.  Cambridge:  National  Bureau  of  Economic  Research. 

Handl,  Johann,  1996:  Hat  sich  die  berufliche  Wertigkeit  der  Bildungsabschlusse  in  den  achtziger 
Jahren  verringert?  Eine  Analyse  der  abhcingig  erwerbstatigen,  deutschen  Berufsanfbnger  auf 
der  Basis  von  Mikrozensusergebnissen.  Kolner  Zeitschrift  fur  Soziologie  und 
Sozialpsychologie  48,  249-273. 

Hannan,  Damian  F.,  David  Raffe,  and  Emer  Smyth,  1997:  Cross-National  Research  on  School  to 
Work  Transitions:  An  Analytical  Framework.  In:  Patrick  Werquin,  Richard  Breen  and  Jordi 
Planas  (eds.):  Youth  Transitions  in  Europe:  Theories  and  Evidence.  Marseille:  CEREQ,  409- 
442. 

Heath,  Anthony  and  Sin  Yi  Cheung  1998:  Education  and  Occupation  in  Britain.  In:  Yossi  Shavit  and 
Walter  Muller  (eds.),  From  School  to  Work.  A Comparative  Study  of  Educational  Qualifications 
and  Occupational  Destinations.  Oxford:  Clarendon  Press. 

Helberger,  Christof,  Ulrich  Rendtel,  and  Johannes  Schwarze,  1994:  Labour  Market  Entry  of  Young 
People  Analysed  by  a Double  Treshold  Model.  In:  Johannes  Schwarze,  Friedrich  Buttler  and 
Gert  G.  Wagner  (eds.):  Labour  Market  Dynamics  in  Present  Day  Germany.  Frankfurt,  Boulder: 
Campus,  Westview,  142-164. 

Jencks,  C.  et  al„  1972:  Inequality:  A Reassessment  of  the  Effect  of  Family  and  Schooling  in  America. 
New  York:  Basic  Books 

Kalleberg,  Arne  L.,  Ivar  Berg,  1987:  Work  and  industry  : Structures,  Markets,  and  Processes.  New 
York:  Plenum,  XVIII,  244. 

Kerckhoff,  Alan  C.,  1996:  Building  Conceptual  and  Empirical  Bridges  Between  Studies  of  Educational 
and  Labour  Force  Careers.  In:  Alan  C.  Kerckhoff  (ed.):  Generating  Social  Stratification: 
Toward  a New  Research  Agenda.  Boulder:  Westview  Press,  37-56. 


O 

ERJC 


1.31 


167 


Kerckhoff,  Alan  C.,  (forthcoming):  Transition  from  School  to  Work  in  Comparative  Perspective.  In: 
Maureen  T.  Hallinan  (e<±),  Handbook  of  Sociology  of  Education,  mimeo. 

Kettunen,  Juha,  1994:  The  Effects  of  Education  on  the  Duration  of  Unemployment.  Labour  8,  331-352. 

Konig,  Wolfgang,  Paul  Luttinger  and  Walter  Muller,  1988:  A Comparative  Analysis  of  the  Development 
and  Structure  of  Educational  Systems.  Methodological  Foundations  and  the  Construction  of  a 
Comparative  Educational  Scale.  CASMIN  Working  Paper  No.  12.  Mannheim:  Institut  fur 
Sozialwissenschaften. 

Lane,  Christel,1992:  Management  and  Labour  in  Europe  : The  Industrial  Enterprise  in  Germany, 
Britain  and  France.  Aldershot:  Edward  Elgar. 

Layard,  Richard,  Stephen  Nickell,  and  Richard  Jackman,  1991:  Unemployment.  Macroeconomic 
Performance  and  the  Labour  Market.  Oxford:  Oxford  University  Press. 

Maurice,  Marc,  Francois  Sellier  and  Jean-Jacques  Silvestre,  1982:  Politique  d’education  et 
organisation  industrielle  en  France  et  en  Allemagne.  Essai  d'analyse  societale.  Paris:  Presses 
Universitaires  de  France 

Marsden,  David,  1990:  Institutions  and  Labour  Mobility.  Occupational  and  Internal  Labour  Markets  in 
Britain,  France,  Italy  and  West  Germany.  In:  Brunette,  Renato  and  C.  DellAringa  (eds.): 
Labour  Relations  and  Economic  Performance.  London:  MacMillan. 

Marsden,  David  and  Paul  Ryan,  1990:  Institutional  Aspects  of  Youth  Employment  and  Training  Policy 
in  Britain.  British  Journal  of  Industrial  Relations  28,  351-369. 

Marsden,  David  and  Paul  Ryan,  1991a:  Initial  Training,  Labour  Market  Structure  and  Public  Policy: 
Intermediate  Skills  in  British  and  German  Industry.  Ryan,  Paul  (ed.)  International  Comparisons 
of  Vocational  Education  and  Training  for  Intermediate  Skills.  London:  Falmer  Press  ,251-285. 

Marsden,  David  and  Paul  Ryan,  1991b:  Institutional  Aspects  of  Youth  Employment  and  Training 
Policy:  Reply.  British  Journal  of  Industrial  Relations  29,  3,  497-505. 

Marsden,  David  and  Paul  Ryan,  1995:  Work,  Labour  Markets  and  Vocational  Preparation:  Anglo- 
German  Comparisons  of  Training  in  Intermediate  Skills.  Bash,  Leslie  and  Andy  Green  (eds.) 
Youth,  Education  and  Work.  World  Yearbook  of  Education  1995.  London:  Kogan  Page,  67-79. 

Mincer,  Jacob,  1994:  Human  Capital:  A Review.  In:  Clark  Kerr  et  al.  (eds.):  Labour  Economics  and 
Industrial  Relations:  Markets  and  Institutions.  Cambridge:  Harvard  University  Press,  109-141. 

Muller,  Walter,  and  Yossi  Shavit,  1998:  The  Institutional  Embeddedness  of  the  Stratification  Process: 
A Comparative  Study  of  Qualifications  and  Occupations  in  Thirteen  Countries.  In:  Yossi  Shavit 
and  Walter  Muller  (eds.):  From  School  to  Work.  A Comparative  Study  of  Educational 
Qualifications  and  Occupational  Destinations.  Oxford:  Clarendon  Press,  1-48. 

OECD,  1996:  Growing  into  Work:  Youth  and  the  Labour  Market  over  the  1980s  and  1990s. 
Employment  Outlook  58,  109-159. 

OECD,  1998:  OECD  Education  at  a Glance.  OECD  Indicators  1998.  Paris:  OECD 

Ruiz-Quintanilla,  Segundo  A.  and  Rita  Claes,  1996:  Determinants  of  Underemployment  of  Young 
Adults:  A Multi-Country  Study.  Industrial  and  Labour  Relations  Review  49,  424-438. 

Shavit,  Yossi  and  Walter  Muller,  1998:  From  School  to  Work.  A Comparative  Study  of  Educational 
Qualifications  and  Occupational  Destinations.  Oxford:  Clarendon  Press. 

Sorge,  Arndt,  1983:  Die  betriebliche  Erzeugung  und  die  Nutzung  beruflicher  Bildung  in  der 
Bundesrepublik  Deutschland,  Frankreich  und  GrolJbritannien.  In:  Haller,  Max  and  Walter 


1.32 


168 


Muller  (eds.):  Beschaftigungssystem  im  gesellschaftlichen  Wandel.  Frankfurt:  Campus,  270- 
286. 

Spence,  Michael,  1981:  Signaling,  Screening,  and  Information.  In:  Rosen,  Sherwin  (ed.):  Studies  in 
Labour  Markets.  Chicago:  University  of  Chicago  Press. 

Spence,  Michael,  1973:  Job  Market  Signaling.  Quarterly  Journal  of  Economics  87,  355-374. 

Spilerman,  Seymour,  1986:  Organisational  Rules  and  the  Features  of  Work  Careers.  Research.  In: 
Treiman,  Donald  and  R.  Robinson  (eds.):  Social  Stratification  and  Mobility.  Greenwich 
(Conn.):  JAI. 

Spilerman,  Seymour  and  Lunde,  1991:  Features  of  Educational  Attainment  and  Job  Promotion 
prospects.  American  Journal  of  Sociology  97,  689-720. 

Statistisches  Bundesamt  (ed.),  1982:  Stand  und  Entwicklung  der  Erwerbstatigkeit.  Fachserie  1: 
Bevolkerung  und  Erwerbstatigkeit,  Reihe  4.1.1.  Kohlhammer,  Stuttgart  und  Mainz. 

Stigler,  G.,  1962:  Information  in  the  Labour  Market.  Journal  of  Political  Economy,  94-105. 

Thurow,  Lester  C.,  1975:  Generating  Inequality.  Mechanisms  of  Distribution  in  the  U.S.  Economy.  New 
York:  Basic  Books. 

Thurow,  Lester  C.,  1978:  Die  Arbeitskrafteschlange  und  das  Modell  des  Arbeitsplatzwettbewerbs.  In: 
Sengenberg,  W.  (ed.):  Der  gespaltene  Arbeitsmarkt.  Probleme  der  Arbeitsmarktsegmentation. 
Frankfurt/Main:  Campus. 

Winkelmann,  Rainer,  1996:  Employment  Prospects  and  Skill  Acquisition  of  Apprenticeship-trained 
Workers  in  Germany.  Industrial  and  Labour  Relations  Review  49,  658-672 


1.33 


169 


Appendix 


Table  A1-1:  Labour  Market  Positions,  based  on  EGP  and  Sectoral  Position 


Labour  Market  Position 

Composition 

Professional 

1,  II 

Skilled 

Ilia,  V,  VI,  lllb  (if  employed  in  public  sector) 

Self-employed 

IVabc,  Selfemployed  form  1,  II, 

Unskilled 

Vllab,  lllb  (if  not  employed  in  public  sector) 

Table  A1-2:  The  EGP  Classification 


Classes 

1 


Ilia 

lllb 

Ivab 

Ivc 

V 

VI 

Vllab 


Description 

Higher-grade  professionals,  administrators,  and  officials;  managers  in  large  industrial 
establishments;  large  proprietors 

Lower  grade  professionals,  administrators,  and  officials;  higher-grade  technicians; 
manager  in  small  industrial  establishment;  supervisors  of  non-manual  employees 

Routine  non-manual  employees,  higher  grade  (in  administration  and  commerce) 

Routine  non-manual  employees,  lower  grade  (sales  and  services) 

Small  proprietors  and  artisans  with  or  without  employees 

Farmers  and  smallholders;  other  self-employed  in  primary  production 

Lower-grade  technicians,  supervisors  of  manual  workers 

Skilled  manual  workers 

Semi-  and  unskilled  manual  workers,  Agricultural  and  other  Workers  in  primary  Industry 


Source:  adapted  from  Brauns/Haun/Steinmann  1997,  S.  5 


170 

\ f 


1.34 


A2:  Early  Career  Unemployment  Risks  in  Three  European  Countries,  Logit  Models 


France 

United  Kingdom 

West  Germany 

1984 

1994 

1984 

1994 

1982 

1995 

Intercept 

-0,60  (.04) 

-0,31  (.05) 

-0,54  (.04) 

-0,28  (.05) 

-1,01  (.05) 

-0,70  (.11) 

Women 

0,46  (.04) 

0,20  (.05) 

-0,29  (.05) 

-0,54  (.05) 

0,25  (.04) 

0,08  (.08) 

n.s. 

Educational  Qualifications 

(REF:  CASMIN  lab) 

- CASMIN  1c 

-0,77  (.06) 

-0,93  (.07) 

-0,79  (.10) 

-0,46  (.08) 

-1,65  (.05) 

-1,91  (.13) 

- CASMIN  2b 

-0,67  (.08) 

-0,52  (.09) 

-0,99  (.05) 

-1,42  (.10) 

-0,68  (.09) 

-0,82  (.17) 

- CASMIN  2a 

-1,13  (.07) 

-1,11  (.08) 

-1,20  (.09) 

-1,37  (.09) 

-2,21  (.07) 

-2,59  (.13) 

- CASMIN  2c 

-1,23  (.10) 

-0,86  (.10) 

-1,22  (.09) 

-1,35  (.10) 

-0,92  (.09) 

-1,74  (.17) 

- CASMIN  2c  voc 

-1,56  (.11) 

-1,05  (.08) 

-1,76  (.17) 

-1,68  (.14) 

-1,74  (.10) 

-2,85  (.17) 

- CASMIN  3a 

-2,39  (.12) 

-1,74  (.08) 

-1,89  (.17) 

-2,01  (.15) 

-1,81  (.12) 

-2,44  (.19) 

- CASMIN  3b 

-2,58  (.14) 

-1,91  (.09) 

-1,93  (.10) 

-1,91  (.10) 

-1,86  (.10) 

-2,16  (.15)  > 

N 

12.961 

11.621 

14.269 

11.178 

33.518 

13.644 

Log-likelihood  Lo 

-6.959,18 

-6.330,49 

-7.056,55 

-5.263,11 

-10.577,50 

-4.129,38 

Log-likelihood  Lt 

-6.343,71 

-5.941,48 

-6564.81 

-4838,40 

-9.866,61 

-3.804,81 

Likelihood  ratio  test  (df) 

1.230,94  (8) 

778,02  (8) 

983,49  (8) 

849,41  (8) 

1.433,64  (8) 

451,11(8) 

R*Ml 

0,09 

0,06 

0,07 

0,08 

0,04 

0,04 

BIC’ 

-1.155,18 

-703,14 

-906,96 

-774,84 

-1.337,84 

-570,74 

Notes: 

Standard  errors  in  parentheses:  n.s . signifies  statistical  significance  at  p > .05; 

Sources: 

Enquete  Emploi  1984  and  1994;  Mikrozensus  1982  and  1995;  UK  Labour  Force  Survey  1984  and  1994; 
Entrants  into  the  labour  force,  unweighted  results 


1.35 


171 


January  2001 


Prepared  as  part  of  the  TSER  project: 

Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe 


The  Position  of  Young  People  and  New  Entrants  in 
European  Labour  Markets 


Thomas  Couppie 
Michele  Mansuy 


Cereq 

10  place  de  al  Joliette 
F- 13474  Marseille  Cedex  02 


France 

E-mail:  couppie@cereq.fr 

michele.mansuy@insee.fr 


o 

ERIC 


172 


0RKING  PAPERS 


Introduction 


The  aim  of  this  chapter  is  to  present  a broad  descriptive  overview  of  patterns  and  labour 
market  outcomes  characterising  the  school-to-work  transition  in  European  Union  countries. 
Numerous  studies  have  already  dealt  with  these  issues,  both  in  national  research  and  in 
comparative  analysis  (e.g.  Shavit  and  Muller  1998,  Hannan,  Raffe  and  Smyth  1997,  Hannan 
and  Werquin  1999,  Hannan  and  alii  1999,  Ryan  1999).  They  single  out  the  large  variety  of 
possible  patterns  and  outcomes  of  the  school-to-work  transition.  Firstly,  they  point  out  the 
possible  linkages  between  the  labour  market  and  the  education  and  training  system.  Secondly, 
they  focus  on  the  labour  market  integration  of  young  inexperienced  people,  by  analysing  job 
access  conditions  and  the  resulting  unemployment  risks  as  well  as  investigating  the  specific 
instability  of  jobs  held  in  the  transition  process.  Thirdly,  they  focus  on  the  job  specificity  of 
young  inexperienced  people.  Their  occupational  attainment  as  well  as  the  nature  of  firms 
hiring  them  appears  to  be  more  or  less  distinctive.  Meanwhile,  the  blurred  nature  of  the 
transition  concept  emerges  as  its  changing  outlines  and  characteristics  across  European 
countries  are  underlined.  As  a result,  no  common  and  comprehensive  definition  of  that 
concept  is  sufficiently  developed  to  pinpoint,  in  a single  and  straightforward  fashion,  key 
events  delimiting  the  timing  of  the  process  and  the  individuals  involved  (Rose  1998,  Vincens 
1997, 1998). 

Although  researchers  do  not  fully  agree  on  a precise  definition  of  the  transition,  they  still 
permit  a loose  identification.  Broadly  speaking,  youth  transition  from  school-to-work  can  be 
summarised  as  the  passage  from  school  to  a relatively  stable  position  in  working  life.  It  is  thus 
a dynamic  process  where  individuals  are  singled  out  by  their  leaving  position  from  the 
education  and  training  system  and  also  by  their  entry  position  on  the  labour  market.  This 
phenomenon  has  implications  at  the  micro  level  and  at  the  macro  level.  At  the  individual 
level,  this  phase  corresponds  to  the  process  of  acquiring  experience  on  the  labour  market. 
Education  and  training  qualifications  are  thus  converted  into  working  abilities  and  job 
positions.  At  the  macro  level,  transition  is  a blurred  period  where  birth  cohorts,  previously 
linked  by  their  collective  participation  in  the  education  and  training  system,  become 
separated.  At  the  two  levels,  the  process  is  present  in  all  the  EU  countries.  But  it  takes  various 
forms,  depending  on  the  characteristics  of  the  national  educational  system  and  their 


0 


2.1 


1 73 


interactions  with  labour  market  organisation.  Thus,  the  biographical  timing  of  the  transition 
process  and  the  associated  forms  of  activity  are  likely  to  differ  from  one  country  to  another. 

The  following  presentation  is  based  on  the  European  Union  Labour  Force  Surveys  of  1995, 
1996  and  1997.  These  surveys  essentially  collect  cross-sectional  data  with  information 
available  by  age1.  In  the  first  part,  we  intend  to  delimit  the  biographical  timing  of  the 
transition  in  order  to  empirically  delineate  its  outlines.  This  leads  us  to  a description  of  some 
cross-national  differences  in  transition  features.  Then,  we  introduce  the  notion  of  new  entrants 
as  a more  useful  category  in  analysing  the  transition  process  than  birth  cohorts.  In  the  third 
part,  we  present  some  major  characteristics  of  LM  outcomes  for  new  entrants.  The  conditions 
of  their  integration  on  the  LM  are  presented  with  a particular  focus  on  the  mobility  issue.  The 
quality  of  jobs  they  held  is  also  investigated.  The  last  part  is  devoted  to  a brief  reminder  of  the 
role  of  educational  attainment  in  transition  outcomes. 


1 The  reader  should  keep  in  mind  that  all  indicators  presented  by  age  group  are  constructed  from  data  collected  on  different 
birth  cohorts.  Thus,  interpreting  what  is  observed  at  the  moment  as  what  is  expected  to  happen  to  a birth  cohort  in  the 
course  of  its  life  is  based  on  the  assumption  of  a certain  permanency  of  institutions,  regulations  and  agents’  behaviours 
on  the  labour  market  and  in  the  education  and  training  system. 


The  ‘youth’  perspective  : biographical  timing  of  the  labour 
market  entry  in  Europe 

Different  patterns  of  ending  participation  in  education  and  training  system  (ETS) 

Compulsory  school  is  a common  feature  across  all  European  countries  and,  although  the 
minimum  age  varies  between  countries,  it  warrants  that  almost  every  young  person  under  14 
attends  some  kind  of  course.  Beyond  that  stage,  participation  in  the  ETS2  follows  a linearly 
declining  trend  as  age  increases  (Figure  1).  Finally  all  national  situations  converge  toward  a 
residual  level  of  participation  in  ETS.  But  the  whole  process  leading  from  a starting  point 
with  everybody  in  ETS  to  an  ending  point  with  a stabilised  residual  level  of  participation  is 
dramatically  differentiated  between  countries. 

National  peculiarities  are  present  in  all  the  main  features  of  the  process.  The  starting  point  and 
the  pattern  of  leaving  ETS  vary  strongly.  The  process  has  already  started  at  15  in  certain 
countries  (Greece,  Italy  and  Portugal)  while  it  has  hardly  begun  at  18  for  others  (Belgium, 
Denmark,  Germany  and  France).  The  pattern  itself  is  far  from  being  regular  between  as  well 
as  within  countries.  Some  countries  seem  to  show  a slow  decrease  in  participation  among  the 
youngest  age  cohorts,  then  a dramatic  fall  to  finish  with  a smooth  slip  down  (e.g.  Belgium). 
Others  show  a quick  fall  that  slows  down  as  age  increases  (United  Kingdom).  For  some 
countries,  the  participation  rate  may  temporarily  halt  its  decline  (e.g.  Austria  or  Sweden).  In 
the  same  way,  the  ending  point  of  the  leaving  process  shows  important  variations.  The  age 
showing  a stabilisation  to  a residual  level  of  participation  in  ETS  (a  maximum  variation  in  the 
participation  rate  of  2%  between  two  successive  age  cohorts)  ranges  from  24  (United 
Kingdom)  to  30  (Italy). 

The  dramatic  variety  across  countries  can  be  underlined  by  the  minimum  and  maximum 
participation  rates  observed  for  every  age  among  the  15  EU  countries.  The  gap  between  the 
two  curves  is  never  lower  than  17%  between  the  15  years-old  age  cohort  and  the  26  years-old 
age  cohort.  It  rises  to  about  40%  for  the  19  years-old  age  cohort,  contrasting  France  (80.5%) 
and  the  United  Kingdom  (42.8%).  Another  way  to  underpin  the  dramatic  differentiation  of 
national  patterns  is  to  single  out  the  age  range  in  the  leaving  process.  The  passage  from  a 90% 


2 Whatever  the  nature  of  ETS  track,  including  lower  secondary 
apprenticeship. 


education  and  training  courses,  higher  education  courses  and 


participation  rate  in  ETS  to  a 10%  rate  ranges  over  8 year  groups  in  Belgium  (from  18  to  25) 
while  it  ranges  over  16  year  groups  in  Italy  (from  15  to  30). 

The  differentiation  of  the  leaving  process  described  above  mainly  has  its  origin  in  the  variety 
of  national  organisation  of  the  ETS.  Leaving  points  correspond  to  the  attainment  of  different 
education  and  training  routes  offered  to  pupils,  students,  trainees  and  apprentices.  However, 
many  reasons  contribute  to  explaining  the  rather  smooth  curves  of  the  participation  rate  that 
are  observed  rather  than  successive  'jumps'  corresponding  to  leaving  points.  Inside  the  ETS, 
the  differentiation  of  tracks  and  curricula  contribute  to  a multiplicity  of  exit  points.  The 
existence  of  linkages  between  tracks  and  curricula  is  another  source  of  age  variation  at  a 
given  leaving  point.  Repeating  a year  is  another  factor  increasing  the  range  of  ages  when 
leaving  ETS. 

Beyond  the  differentiation  of  routes  within  the  ETS,  the  issue  of  possible  prospects  open  to 
ETS  participants  is  the  next  issue  to  be  dealt  with.  A first  clear  alternative  is  offered  to  young 
people:  entering  the  labour  market  as  part  of  the  labour  force  or  staying  out  of  it. 


A linear  rise  in  the  activity  rate  among  youth 

Between  the  ages  of  15  and  30,  most  of  the  young  choose  to  enter  the  labour  market.  This 
decision  may  occur  during  the  course  of  their  studies.  But  more  often,  it  happens  after  they 
have  left  education  and  training  programmes.  Indeed,  through  age  cohorts,  the  activity  rate 
rises  as  participation  in  education/training  declines.  For  young  school  leavers,  staying  out  of 
the  labour  force  is  a marginal  behaviour  as  indicated  in  figure  2.  The  proportion  of  people 
who  are  inactive  apart  from  educational  participation  remains  low,  although  it  increases 
gently  but  continuously  with  age. 

Beyond  these  common  trends,  diversity  between  European  countries  is  even  wider  than  for 
training  participation.  The  minimum  and  maximum  activity  rates  observed  for  every  age 
among  the  15  EU  countries  show  wide  gaps  (around  20%  or  more)  through  15  to  25  with  a 
maximum  interval  of  about  70%  at  17  contrasting  Belgium  (3.1%)  and  Denmark  (73.0%). 
Again,  the  dramatic  differences  in  national  behaviour  are  underpinned  by  the  age  range  in  the 
activity  increase.  The  passage  from  a 10%  activity  rate  to  a 75%  rate  ranges  over  5 year 


groups  in  Denmark  (from  15  to  19  years  old)  while  it  ranges  over  16  year  groups  in  Italy 
(from  15  to  30  years  old). 

So,  three  profiles  of  the  development  of  labour  force  participation  can  be  briefly  identified. 
The  first  profile  aggregates  countries  where  there  is  an  important  association  between  an 
increase  in  activity  and  a decline  in  training.  The  second  profile  is  quite  similar  to  the  first  but 
the  increase  in  activity  is  tempered  by  the  early  role  of  inactivity  not  associated  with  ET 
participation  (Greece,  Italy  and  partly  Sweden).  The  third  profile  is  rather  different.  Countries 
belonging  to  it  show  dramatic  increases  in  activity  before  the  fall  of  training  participation. 
The  explanation  lies  in  the  existence  of  situations  where  training  participation  is  combined 
with  having  a job.  These  combined  situations  have  different  origins  and  intensity  in  the  EU 
countries  but  they  all  share  the  common  feature  of  being  transitory  events  related  specifically 
to  young  people. 


Specific  transitory  intermediate  statuses:  the  combination  of  working  and  training. 

A first  specific  feature  of  youth  participation  in  the  labour  market  consists  in  borderline 
situations  bordering  on  one  side  the  trainee,  pupil  or  student  status  and  on  the  other  side  the 
worker  status.  These  situations  imply  simultaneous  participation  in  the  two  activities,  albeit 
patterns  of  combinations  vary  strongly.  The  scope  of  combinations  can  be  ordered  according 
to  the  respective  importance  of  each  activity  to  the  young.  It  ranges  from  a school-dominant 
position  associated  with  a working  student  profile  to  a working-dominant  position  associated 
with  a training  worker  profile.  Extreme  positions  can  be  illustrated  by  the  secondary  role  of 
the  minor  activity.  In  the  first  case,  working  tends  to  be  a subsidiary  function  in  order  to 
ensure  the  continuity  of  the  training  process.  In  the  latter  case,  training  tends  to  be  an 
enhancement  pattern  of  job  skills  scheduled  by  the  employer.  Between  these  positions, 
apprenticeship  stands  out  as  a more  balanced  profile  that  links  training  and  working  activities 
as  joint  elements  of  qualification  production. 

Practically,  however,  grouping  the  combined  statuses  according  to  the  link  between  the  two 
activities  can  be  hard  to  establish  in  surveys  like  LFS.  Nevertheless,  the  overall  importance  of 
those  combined  situations  reveals  that  the  European  Union  can  be  split  into  two  groups  of 
countries.  The  first  group  aggregates  countries  where  these  situations  have  a rather  weak 


2.5 


177 


impact  on  the  pattern  of  transition  (figure  3).  In  these  countries,  the  combined  status  of 
worker  and  trainee  is  infrequent,  never  exceeding  10%  for  an  age  group.  The  link  between 
training  and  working  can  be  viewed  mainly  as  a turning  point  between  the  two  activities.  The 
second  group  aggregates  countries  where  those  situations  have  a more  extensive  impact  on 
the  pattern  of  transition.  The  set  of  countries  is  more  restricted.  It  includes  Denmark, 
Germany,  Netherlands,  Austria,  Sweden  and  United  Kingdom.  The  importance  of  combined 
status  rises  to  20%  for  some  age  groups  and  can  reach  far  higher  levels.  The  nature  of  such 
situations  is  transitory,  as  shown  by  the  limited  range  of  age  groups  involved.  This  range  is  at 
its  greatest  in  Denmark  with  10  year  groups  (15  to  24)  being  above  the  20%-share  of 
combined  status  among  them.  However,  that  doesn’t  mean  these  countries  have  close  patterns 
of  combining  the  two  activities.  Germany  and  Austria,  with  well-developed  apprenticeship 
organisation,  can  be  contrasted  to  the  Netherlands  and  Sweden  with  dominant  non  co- 
ordinated combinations  of  training  and  working.  Both  types  of  combinations  are  present  in 
Denmark  and  United  Kingdom. 

Combined  training  and  working  situations  represent  a first  specific  feature  of  the  youth 
transition  from  school  to  work.  This  feature  is  not  the  only  one  to  be  considered,  as  entering 
the  labour  market  does  not  necessarily  mean  access  to  a job.  Unemployment  can  be  viewed  as 
another  transitory  situation,  although  it  also  concerns  experienced  workers. 


Access  to  jobs:  unemployment  as  an  initial  rough  measure  of  difficulties  in  entering 
companies 

The  ultimate  aim  of  young  people  in  the  transition  period  is  to  obtain  a good  employment 
position.  This  process  can  in  some  cases  start  -and  succeed-  before  the  end  of  schooling,  but  it 
usually  extends  beyond  this  time.  Unemployment  simply  describes  the  situation  of  young 
people  seeking  a job.  This  is  not  a situation  specific  to  young  people  but,  in  their  case,  it 
assumes  a specific  level  and  reaches  a momentum  that  singles  out  young  people  from  other 
labour  force  groups.  Moreover,  the  nature  of  this  momentum  is  essentially  transitory,  as  is  the 
case  observed  for  intermediate  statuses  combining  training  and  working. 

There  are  two  ways  to  measure  the  importance  of  unemployment  among  different  age  groups. 
First,  the  unemployment  rate  can  be  interpreted  as  a risk  indicator  as  it  compares  the 

o • r « 

ERIC  ” ' ' 


*•«  178 


unemployed  to  the  total  labour  force.  Second,  the  unemployment  proportion  can  be 
interpreted  as  an  extent  indicator  as  it  relates  the  unemployed  to  the  whole  population.  The 
first  evaluates  the  intensity  of  difficulties  on  the  labour  market  while  the  second  measures  the 
global  impact  of  unemployment  among  different  age  groups3.  The  way  the  two  indicators 
combine  together  leads  to  different  macro  economic  interpretations  of  young  people's  LM 
position.  For  example,  a proportionately  high  level  of  unemployment  indicates  that  a large 
group  of  young  people  is  facing  difficulties  in  accessing  jobs  while  a significant  difference 
between  the  unemployment  proportion  and  rate  indicates  enhanced  risks  to  a restricted 
subgroup. 

Figure  4 shows  the  two  indicators’  evolution  among  age  groups  and  also  depicts  EU 
minimum  and  maximum  points  of  unemployment  proportions.  No  common  tendency  across 
Europe  can  be  established.  On  the  one  hand,  quasi-constant  and  low  levels  of  youth 
unemployment  can  be  found  in  Denmark,  Germany,  Austria  and  to  a certain  extent  the 
Netherlands.  The  two  measures  remain  close  at  every  age  over  17.  In  these  countries,  young 
people  do  not  seem  to  have  specific  difficulties  in  accessing  and  remaining  in  jobs.  On  the 
other  hand,  the  early  increase  and  later  progressive  fall  of  the  unemployment  proportion 
indicate  a transitory  regime  applying  to  youth  unemployment.  The  peak  and  the  spread  of 
such  a regime  vary  between  countries.  It  is  particularly  noticeable  in  some  countries  (Greece, 
Spain,  Italy,  France  and  Finland),  less  pronounced  in  others  (Ireland,  Sweden,  Portugal  and 
United  Kingdom).  The  most  extreme  situation  applies  to  Spain  with  unemployment 
proportions  peaking  at  25  years  old  and  still  declining  after  the  age  of  30.  This  feature 
indicates  widespread  difficulties  throughout  youth  in  accessing  jobs.  Meanwhile,  the 
unemployment  rate  reaches  high  levels  early  with  a peak  between  15  and  20  years-old, 
topping  60%  or  more  of  young  actives  in  some  countries  (Finland,  Spain).  Then  it  rapidly 
decreases  to  converge  towards  the  unemployment  proportion.  This  feature  highlights 
transitions  occurring  at  early  ages  as  highly  risky  events  and  points  out  early  school  leavers  as 
a disadvantaged  category  on  the  labour  market.  Among  countries  fitting  this  profile,  United 
Kingdom  is  a borderline  case  as  it  shows  a limited  extent  of  unemployment,  a slight  decline 
as  age  increases  and  a rapid  reduction  in  unemployment  risk. 


3 The  unemployment  rate  is  always  higher  than  or  equal  to  the  unemployment  proportion  and  the  two  tend  to  converge  as  the 
activity  rate  rises.  They  become  equal  among  groups  when  everybody  is  active. 

r 179 


2.7 


Having  detailed  the  main  possible  positions  for  young  people  in  and  out  of  the  labour  market, 
a brief  summary  of  national  situations  would  help  to  highlight  some  national  specific 
characteristics  of  school-to-work  transition. 


Figure  1:  Proportion  of  people  attending  an  education 
or  training  programme  - average  1995-1997 


Figure  la  : “OLM-type”  countries 


\ r 


2.8 


Figure  lb:  Southern  countries 


100% 

^ EL 

90% 

\ 

80% 

♦ • A ' * 

A 

7 0 % 

x V. 

60% 

\ \ 

50% 

X v 

v X x 

40% 

• “V 

30% 

^ X 

20% 

^ . \ N 

x ^ V N 

1 0 % 

0% 

tS  16  It  U 19  20  21  22  23  24  25  2«  22  25  29  10  11  15  >39 

100%. 

90% 

IT 

80  % 

A i 

70% 

\ 

\ 

60% 

\ v 

50% 

\ X *%  % 

40% 

• A.  s 

3 0 % 

20% 

ONL  V 

> "S. 

1 0 % 

**  - >* 

n h. 

1 5 1 5 1 7 1 1 1 1 20  21  22  2)  24  25  29  27  21  29  50  33  35  >39 

181 


2.9 


Figure  lc:  Other  countries 


100% 

90% 

60% 

70% 

60% 

50% 

40% 

30% 

20% 

10% 

0% 

FR 

\ \v 

\ *<. 

*.  _ -N.  ""  N 

IS  It  17  II  II  20  21  22  21  2*  2S  21  27  21  21  10  11  IS  >11 

i no*  . 

90% 
6 0% 

LU 

7 0% 

• %%  \ 

60  % 

\\y 

50  % 

\ \v* 

40% 

7 — V 

30  % 

*'  \ 

20  % 

; ^ 

1 0%  - 

* * ^ 

0% 

* m 7**  * 

IS  11  17  11  it  20  21  22  21  2«  2S  21  27  21  21  10  11  IS  >11 

100% 

90% 

VjV  SE 

6 0 % 

70% 

' \ V' 

60% 

\ ^ 

50% 

\ ^ 

40% 

\ s ^ 

30% 

20% 

10% 

■ - - _ 

IS  tl  17  11  It  20  21  22  22  24  2S  21  27  21  2*  30  11  IS  >11 

Figure  Id:  European  Union  (EU-15) 


182 


O 


Figure  2:  Proportions  of  active  and  inactive  people,  by  age  - average  1995-1997 


Figure  2a  : “OLM-type”  countries 


100% 
9 0 % 

DK  ^ 

8 0 % 

70% 

A t/ n , 

6 0 % 

5 0 % 

/ 

4 0%. 

/ / 

30% 
2 0 % 

* 7 

10% 

0% 

XxXXlH<X>H()(aC 

1*  te  17  II  19  20  2 1 22  21  24  21  21  27  21  21  10  11  11  >11 


11  It  <7  II  II  20  21  22  21  24  21  21  27  21  21  10  11  11  >11 


Figure  2b:  Southern  countries 


100%- 

90%  - 

EL 

80%  - 

70%  - 

... 

60%  - 

Jf  ✓ 

50%  - 

WT 

40% 

f' 

30%  - 

/✓ 

20%  - 

10% 

0% 

<1  It  1 7 1 1 1 1 20  2 1 22  21  24  21  21  27  21  21  10  11  11  >11 

100% 

90% 
80%  - 

IT 

” 

70% 

60% 

50% 

40% 

y7 

30% 

yV 

20% 

/ * 1 1 >f  tf’’* 

10% 

A 'l  H 

X-*r-x*£ 

0% 

<5  tt  tr  1|  1 9 20  21  22  21  24  21  2t  27  21  21  10  11  11  »16 

2.11 

r > i 
s 

. . -• 


183 


Figure  2c:  Other  countries 


2.12 


184 


Figure  3:  Situations  combining  employment  and  training  by  age  - average  1995-1997 


Figure  3a  : “OLM-type”  countries 


Figure  3b:  Southern  countries 


O 

ER[C 


2.13 


185 


Figure  3c:  Other  countries 


is  IS  17  IS  It  It  21  22  21  24  2S  21  2/  28  2t  10  11  IS  >18 


100% 
9 0 % 

FR 

60  % 

70% 
80  % 

50  % 

T 

/ 

/ 

4 0%. 

! \ 

30% 

7 V'v 

20  % 

*»• 

10% 

a a i 

0% 

IS  18  IT  1 8 1 1 20  21  22  21  24  2S  11  27  21  21  10  11  IS  >11 

100% 

90% 

80% 

70% 

60% 

50% 

40% 

30% 

20% 

10% 

0% 

LU 

r 

• \ 

\ 



* ^ AOS 

^ 0 » » • , . . . . ~ ' — r* ' — v 

is  1 1 17  1 1 1 8 20  21  22  21  24  2S  21  27  28  21  10  11  IS  >11 

inn* 

90  % 

SE 

60  % 

70% 

60% 

50% 

4 0%. 

• 

30% 

i 's 

20% 

10% 

0% 

IS  18  17  II  It  20  21  22  21  24  2S  21  27  28  29  10  11  IS  >11 


Figure  3d:  European  Union  (EU-15) 

100% 

90% 

60% 

70% 

60% 

50% 

40% 

30% 

20% 

10% 

0% 

1 5 1 8 1 7 1 8 1 9 20  21  22  23  24  25  28  27  28  29  30  33  35  >38 


2.14 


Figure  4:  Proportion  of  unemployed  and  unemployment  rate  by  age  - average  1995-1997 


Figure  4a  : “OLM-type”  countries 


70% 

60% 

DE 

50% 

4 0% 

30% 

20% 

/*  ^ 

10% 

v-  " 

0% 

IS  1 1 1 7 1 1 11  20  21  22  21  24  2S  21  27  21  21  10  11  IS  >11 

70% 

6 0% 

AT 

50% 

40% 

30% 

20% 

10% 

yr  ~ ' 

- w \ 

s 

0% 

IS  1 1 1 7 1 1 1 1 20  21  22  21  24  2$  21  27  21  21  10  11  1$  >11 

Figure  4b:  Southern  countries 


IS  <«  <7  IS  IS  70  21  22  21  24  25  26  27  21  21  10  11  IS  >11 


IS  IS  17  II  II  20  21  22  21  24  2S  21  27  21  21  10  11  IS  >11 


70% 


6 0% 

ES 

50% 

V)(.. 

40% 

30% 

20% 

10% 

0% 

s . 

IS  11  17  11  l|  20  21  22  21  24  2$  21  27  21  21  10  11  1$  >11 

70% 

PT 

60% 

50% 

40% 

30% 

20% 

10% 

0% 

is  1 1 1 7 1 1 1 9 20  21  22  71  24  2S  21  27  21  21  10  11  1$  >11 

2.15 

",  f' 

■ * * ' i 


187 


Figure  4c:  Other  countries 


Figure  4d:  European  Union  (EU-15) 


hi 


2.16 


188 


Figure  5:  Biographical  timing  of  the  school-to-work  transition  - Recapitulation  of 

national  features,  average  1995-1997 

Figure  5a  : “OLM-type”  countries 


DENM  ARK 

GERMANY 

illteP— 1„„„ 

Ml 

15  16  17  IB  19  20  21  22  23  24  25  26  27  2S  29  30  33  35  >35 

tge 

IS  16  17  IB  19  20  21  22  23  24  25  26  27  28  29  30  33  35  >35 

age 

NETHERLANDS 

A US  TRIA 

IS  16  17  IB  19  20  21  22  23  24  25  26  27  28  29  30  33  35  >35 

age 

IS  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  33  35  >35 

age 

Figure  5b  : Southern  countries 


GR  EEC E 

S P A IN 

IS  16  17  IB  19  20  21  22  23  24  25  26  27  28  29  30  33  35  >35 

age 

1 5 16  17  18  1 9 20  21  22  23  24  25  26  27  28  29  30  33  35  >35 

age 

IT  A LY 

P ORTUGA  L 

15  16  17  IB  19  20  21  22  23  24  25  26  27  2B  29  30  33  35  >35 

*g* 

IS  16  17  IB  19  20  21  22  23  24  25  26  27  28  29  30  33  35  >35 

•g* 

2.17 


189 


Figure  5c:  Other  countries 


B ELGIUM 

FRANCE 

15  16  1 7 18  19  20  21  22  23  24  25  26  27  28  29  30  33  35  >35 

age 

1 5 16  1 7 1 8 19  20  21  22  23  24  25  26  27  28  29  30  33  35  >35 

»g£ 

IRELAND 

LUX  E M BOURO 

15  16  17  18  19  20  21  22  23  24  25  26  27  28  29  30  33  35  >35 

1 5 16  17  1 8 19  20  2 1 22  23  24  25  26  27  28  29  30  33  3 5 >35 

»g£ 

age 

FINLA  ND 

SWEDEN 

1 5 16  17  1 8 1 9 20  21  22  23  24  25  26  27  28  29  30  33  35  >35 

»g£ 

15  16  17  18  1 9 20  21  22  23  24  25  26  27  28  29  30  33  35  >35 

»g£ 

UNITED  KINGDOM 


IS  16  17  IB  19  20  21  22  23  24  2S  26  27  28  29  30  33  35  >35 

»gt 


Education  Sc  tra  ining  wit  ho  ut  e m plo  ym  c nt  ^Employments  training 
^Unemployment  | ha  c tiv  ity  wit  ho  ut  e due  a tio  n Sc  tra  ining 


■ Em  plo  ym  e nt  wit  ho  ut  tra  ining 


f;  f:  • 

2.18 


1 Qn 


Age  of  youth  entry  into  the  labour  market:  a summary 

The  preceding  results  can  be  summarised  by  country,  as  shown  in  figure  5.  Clearly,  some 
country  profiles  emerge.  They  partly  fit  with  our  formal  clustering  contrasting  ‘OLM-type’ 
countries,  southern  countries  and  the  other  countries. 


‘OLM-type’  countries  share  a quite  similar  profile,  which  contrasts  significantly  with  the 
profiles  of  other  countries  of  the  EU.  The  process  of  entering  the  LM  starts  around  the  age  of 
15.  It  mainly  takes  the  shape  of  combinations  of  training  and  employment  between  15  and  20. 
Unemployment  has  a limited  impact  during  the  whole  process.  However,  other  features  are 
more  national-specific. 

Denmark  is  the  country  where  double  statuses  of  training  and  employment  are  the  most 
developed.  Their  impact  is  quite  concentrated  on  younger  cohorts  but  is  still  significant  after 
the  age  of  25.  Apprenticeship,  although  well  developed,  is  far  from  being  the  only  way  of 
combining  training  and  working.  By  contrast,  employment  without  training  increases 
smoothly  across  age  cohorts.  The  stabilisation  of  positions  on  and  out  of  the  labour  market 
occurs  at  a late  stage,  around  30. 

The  Netherlands  shares  with  Denmark  a large  spread  of  double  statuses  across  age  cohorts. 
These  situations  are  still  evident  after  30.  But,  unlike  in  the  other  ‘OLM-type’  countries, 
apprenticeship  plays  a minor  role  in  combinations  of  training  and  working.  The  growth  of 
employment  without  training  is  more  sustained.  The  stabilisation  of  positions  on  and  out  of 
the  labour  market  occurs  a bit  earlier,  around  27. 

Germany  shows  a strong  but  more  limited  impact  of  double  statuses  across  age  cohorts. 
These  situations  correspond  mainly  to  the  famous  German  ‘dual  system’  based  on 
apprenticeship.  Unlike  Denmark  and  the  Netherlands,  the  young  attending  school-based  ET 
courses  stay  out  of  the  LM.  The  extension  of  employment  without  training  is  continuous  and 
still  in  progress  around  the  age  of  30.  This  can  be  connected  with  some  long  lasting 
participation  in  the  ETS  beyond  the  age  of  25.  Thus,  the  stabilisation  of  positions  on  and  out 
of  the  labour  market  occurs  at  a late  stage,  around  30. 


2.19 


191 


Austria  differs  a bit  more  from  the  three  others  as  double  statuses  have  a more  restricted 
impact,  albeit  important  between  15  and  19.  Like  in  Germany,  these  situations  are  also  mainly 
related  to  apprenticeship.  By  contrast  with  the  other  countries,  the  growth  of  employment 
without  training  is  concentrated  between  17  and  21.  A minor  part  of  long  lasting  participation 
to  ETS  results  in  a rather  late  stabilisation  of  positions  on  and  out  of  the  labour  market, 
around  28. 


Southern  countries  appear  a bit  more  heterogeneous  when  compared  with  the  preceding 
group.  While  Greece,  Spain  and  Italy  share  common  features  leading  to  a rather  specific 
group  profile,  the  case  of  Portugal  differs  strongly.  The  proximity  of  Greece,  Spain  and  Italy 
rests  on  four  characteristics.  First,  unemployment  has  a significant  impact,  covering  numerous 
age  cohorts  between  1 8 and  30.  Second,  combinations  of  training  and  working  are  absent  in 
these  countries.  Third,  the  growth  of  employment  rate  is  smooth  and  lasts  until  30.  Fourth, 
there  is  a progressive  rise  of  inactivity  not  related  to  training  through  age  cohorts.  Still,  the 
three  countries  have  national-specific  features. 

Greece  has  the  least  marked  profile  of  the  three.  Participation  in  ETS  is  strong  until  20,  levels 
of  unemployment  remain  beyond  those  of  Italy  and  Spain  and  employment  rates  make  major 
progress  before  25.  The  stabilisation  of  positions  on  and  out  of  the  labour  market  occurs 
around  28. 

Spain , by  contrast,  is  singled  out  by  its  high  levels  of  unemployment,  peaking  at  25. 
Compared  to  the  others,  inactivity  without  training  has  a more  restricted  impact.  Spain  is  the 
country  of  the  European  Union  where  employment  is  generally  the  least  developed 
throughout  age  cohorts.  Although  employment  reaches  a peak  at  the  age  of  28,  the  decline  of 
unemployment  is  then  still  in  progress  as  inactivity  rises.  Thus,  the  stabilisation  of  positions 
on  and  out  of  the  labour  market  occurs  after  30. 

Italy  has  a midway  position  in  many  aspects  of  its  characteristics  of  unemployment  and 
employment.  It  contrasts  with  the  two  others  in  the  relatively  high  impact  of  inactivity 
through  age  cohorts.  Inactivity  without  training  has  an  early,  significant  influence.  On  the 
other  hand,  ETS  participation  is  still  noticeable  between  25  and  30.  The  stabilisation  of 
positions  on  and  out  of  the  labour  market  occurs  at  a late  stage  in  Italy,  after  30. 


Portugal  presents  a quite  different  profile  to  the  three  former  countries.  Unemployment  is 
rather  moderate  and  without  much  variation  across  age  cohorts.  Although  double  statuses 
haven’t  a strong  influence,  employment  without  training  peaks  at  a higher  level  than  in  other 
southern  countries.  The  stabilisation  of  positions  on  and  out  of  the  labour  market  occurs 
around  28. 


The  other  countries  group  appears  to  be  the  most  heterogeneous  of  the  three.  There  is  no  real 
common  profile  shared  by  all  countries.  Two  countries,  United  Kingdom  and  Finland,  can  be 
partly  linked  with  preceding  profiles  while  France  and  Belgium  on  one  side,  Ireland  and 
Luxembourg  on  another  side,  present  similarities.  Sweden  stays  apart  in  an  intermediate 
position. 

The  United  Kingdom  presents  some  similarities  with  the  ‘OLM-type’  countries.  The  process 
of  entering  the  LM  is  already  in  progress  at  15.  Between  16  and  20,  it  is  also  based  on 
combinations  of  training  and  working,  mainly  by  way  of  Youth  Training.  As  in  Austria, 
employment  without  training  shows  a sustained  growth  before  20.  But  the  British  situation 
differs  in  various  ways.  First,  a more  influential  role  of  unemployment  for  the  youngest  age 
cohorts  can  be  noticed.  Second,  participation  in  the  ETS  sharply  declines  between  20  and  25. 
In  contrast,  employment  without  training  continues  its  vigorous  growth  until  25.  Thus,  the 
whole  process  leading  to  stabilisation  of  positions  on  and  out  of  the  labour  market  through 
age  cohorts  is  achieved  around  25. 

Finland  can  be  partly  linked  with  the  group  of  the  three  southern  countries.  First,  it  shares 
with  them  high  levels  of  unemployment  for  all  age  cohorts  from  15  to  30.  Second,  the  rise  of 
employment  without  training  is  rather  smooth  and  still  in  progress  after  30.  But  Finland 
differs  from  the  others  in  the  more  limited  extent  of  inactivity  without  training,  more 
developed  double  statuses  and  a higher  final  level  of  employment.  As  in  the  southern 
countries,  the  stabilisation  of  positions  on  and  out  of  the  labour  market  occurs  at  a late  stage, 
after  30. 

Belgium  and  France  have  in  common  a late  start  of  the  transition  process.  It  hardly  begins  at 
the  age  of  18  and,  until  the  age  of  20,  a major  part  of  young  people  are  still  in  the  ETS.  Then, 
ETS  participation  dramatically  falls  to  become  residual  after  25.  In  the  two  countries, 


O 

ERIC 


2.21 


unemployment  has  an  influential  role  through  age  cohorts  after  20.  The  impact  of 
combinations  of  training  and  working  is  rather  limited.  Almost  absent  in  Belgium,  double 
statuses  are  a bit  more  extended  in  France  due  to  apprenticeship  and  public  employment 
policies.  The  two  countries  also  differ  in  the  pattern  of  employment  growth,  being  more 
sustained  in  Belgium.  The  stabilisation  of  positions  on  and  out  of  the  labour  market  occurs 
around  27  in  Belgium  while  it  occurs  a bit  later  in  France,  around  30. 

Ireland  and  Luxembourg  have  some  proximity  with  Portugal.  Unemployment  is  rather 
moderate  while  double  statuses  have  a restricted  influence  through  age  cohorts.  Still,  Ireland 
presents  some  peculiarities.  ETS  participation  declines  sharply  until  25  to  become  residual. 
On  the  other  side,  employment  without  training  rapidly  grows  to  peak  at  25  while 
unemployment  tends  to  diminish  slightly.  The  evolution  of  inactivity  without  training  also 
differentiates  Ireland  as  it  grows  gently  across  age  cohorts.  As  in  United  Kingdom,  the 
stabilisation  of  positions  on  and  out  of  the  labour  market  occurs  rather  early,  around  25. 

Luxembourg  is  characterised  by  very  low  levels  of  unemployment.  By  contrast,  as  in 
Portugal,  the  level  of  employment  becomes  high  but  it  occurs  earlier  in  Luxembourg.  The 
stabilisation  of  positions  on  and  out  of  the  labour  market  occurs  around  27. 

The  Swedish  profile  is  characterised  by  the  somewhat  influential  role  of  unemployment  and 
inactivity  without  training  between  20  and  25.  Double  statuses  are  present  but  infrequent. 
There  is  no  apprenticeship  in  Sweden  and  these  cases  are  representative  of  uncoordinated 
training  and  working  combinations.  Unemployment  progresses  smoothly  across  age  cohorts 
while  long  lasting  participation  in  the  ETS  is  also  present.  The  stabilisation  of  positions  on 
and  out  of  the  labour  market  occurs  around  28. 


Changing  perspective:  from  youth  to  new  entrants 


The  previous  analysis  has  allowed  us  to  assess  the  temporal  aspect  of  the  event  and  some 
specific  features  of  the  situations  encountered: 


■ t £ 


2.22 


194 


- The  gradual  transition  of  a birth  cohort  from  education  and  training  to  employment  occurs 
at  variable  ages  and  variable  speeds  depending  on  the  countries.  The  biographical  period  of 
the  transition  begins  at  the  time  the  first  members  of  a birth  cohort  leave  the  education  and 
training  system  and  ends  at  the  time  the  birth  cohort's  participation  in  the  employment  system 
reaches  its  peak.  In  that  sense,  countries  are  highly  differentiated  with  some  of  them 
characterised  by  a rather  early  and  short  period  of  transition  (Austria,  Ireland,  United 
Kingdom)  while  others  are  singled  out  by  a late  start  (France,  Belgium)  or  an  extended  time 
(Italy,  Finland). 

- The  transition  of  birth  cohorts  is  marked  by  numerous  intermediate  situations  between  full- 
time training  and  stability  on  the  labour  market.  Two  types  have  been  singled  out:  situations 
associating  training  with  a work  activity  on  the  one  hand  and  situations  of  unemployment  on 
the  other.  These  situations  are  not  necessarily  specific  to  the  period  of  labour  market  entry, 
but  they  may  be  over-represented  in  that  period.  We  have  already  observed  various 
combinations  of  those  elements  in  the  different  countries. 

The  national  features  of  the  transition  process  turn  out  to  be  extremely  heterogeneous.  They 
raise  the  issue  of  the  influence  of  national  institutions  and  arrangements  on  the  transition 
process.  Obviously,  national  organisation  of  the  ETS  influences  the  biographical  timing  of  the 
transition.  The  existence  of  various  possible  tracks,  the  country-specific  definition  of  degree 
courses  and  the  multiplicity  of  leaving  points  they  generate  inside  the  system  contribute  to  the 
establishment  of  nationally  favoured  ages  for  ending  ETS  participation.  National  LM 
organisation  has  an  impact  too.  The  possible  combinations  of  training  and  working  depend  on 
the  nature  of  LM  arrangements  prevailing  in  the  different  countries.  As  a consequence,  it 
appears  that  strong  links  between  the  ETS  and  the  LM  favour  early  activity.  One  can  also 
relate  the  impact  of  unemployment  in  youth  transition  to  LM  regulation.  The  existence  and 
the  extent  of  specific  youth  unemployment  can  be  interpreted  as  different  national-based 
patterns  of  considering  short  LM  experience  in  hiring  decisions. 

Thus,  the  issue  of  national  institutions  leads  us  to  single  out  some  limitations  of  a comparative 
analysis  of  the  school-to-work  transition  based  on  age  cohorts.  Unable  to  catch  the  specificity 
of  national  institutions,  such  an  analysis  simply  cannot  take  them  into  account  and  therefore 
neglects  their  influence. 


195 


2.23 


As  a result,  it  becomes  difficult  to  further  develop  comparative  analysis  of  the  respective  role 
of  training  attainment  and  LM  experience  during  the  transition  process  based  on  age  cohorts. 
An  alternative  strategy  should  then  be  envisioned  for  further  analysis.  It  should  attempt  to 
deal  with  institutional  factors  resulting  in  differentiation  between  countries.  More 
specifically: 

• The  national  institutional  features  that  influence  the  phasing  of  labour  market  entry  should 
be  integrated.  This  leads  us  to  favour  categories  of  individuals  based  on  the  timing  of  their 
LM  entry  and  the  nature  of  their  school  attainment. 

• The  elements  of  the  transition  process  directly  deriving  from  the  educational  structures 
should  be  separated  from  those  depending  only  on  the  modus  operandi  of  the  labour 
market.  In  order  to  do  so,  it  is  necessary  to  distinguish,  amongst  the  active  individuals, 
those  who  have  completed  their  training  from  those  who  are  pursuing  a training 
programme  in  the  educational  system.  This  serves  to  separate  labour  market  events  that 
can  be  linked,  directly  or  indirectly,  to  a public  educational  policy  regulated  by  institutions 
outside  the  labour  market  from  labour  market  events  resulting  from  the  organisation  and 
modus  operandi  of  the  labour  market  itself. 

On  the  basis  of  these  remarks,  an  alternative  category  of  new  entrants  can  be  useful  to 
improve  the  comparability  of  national  profiles  of  transition.  It  combines  characteristics  of 
their  position  towards  the  education  and  training  system  with  the  experience  accumulated  on 
the  labour  market  (see  appendix  A for  the  construction  of  the  category). 


Labour  market  outcomes  for  new  entrants 


Integration  of  new  entrants  into  the  LM:  job  access  conditions  and  labour 
market  mobility 

A description  of  new  entrants’  integration  into  the  labour  market  follows.  First,  job  access 
conditions  and  unemployment  risks  will  be  developed.  Then,  the  effective  stabilisation  in  jobs 


O 

ERIC 


2.24 


196 


finally  held  is  overviewed.  Finally,  possible  specific  work  conditions  are  investigated  as  a 
potential  source  of  new  entrants’  higher  mobility. 

Figure  6:  Unemployment  rate  according  to  duration  of  the  current  spell,  by  years  of  experience 

on  labour  market 


Figure  6a:  “OLM-type”  countries 


80 
70 
60 
50 
40 
30 
20 
1 0 
0 


D K 95  -97 


80 
70 
60 
50 
40 
30 
20 
1 0 
0 


D E 9 5 • 9 7 


19  11  12  IS  14  1$ 


Figure  6b:  Southern  countries 


B Unemployed  for  more  than  one  year  (%  among  active  people) 
Unemployed  for  less  than  1 year  (%  among  active  people) 


ERIC 


2.25 


197 


Figure  6c:  Other  countries 


Unemployed  for  more  than  one  year  (%  among  active  people) 
Unemployed  for  less  than  1 year  (%  among  active  people) 


o 

ERIC 


2.26 


Unemployment  among  new  entrants 


The  analysis  of  youth  unemployment  has  revealed  two  profiles:  one  with  unspecific  constant 
unemployment  and  the  other  characterised  by  a transitory  regime  somewhat  focused  on 
transitions  occurring  at  early  ages.  The  analysis  of  new  entrants’  unemployment  partially 
confirms  such  assessments. 

Figure  6 recapitulates  unemployment  rate  by  accumulated  years  of  labour  market  experience. 
Of  the  two  previous  regimes  observed  for  youth  unemployment,  the  first  one  tends  to 
disappear  as  Germany  and  Austria  show  a slight  but  consistent  declining  trend  as  experience 
grows.  Still,  these  countries  differ  strongly  from  some  others  in  terms  of  the  limited  extent  of 
unemployment.  For  the  least  experienced  new  entrants,  the  spread  of  the  unemployment  rate 
is  wider;  it  ranges  from  50%  in  several  countries  to  about  10%  in  Luxembourg.  But  for  a 
similar  unemployment  rate,  the  situation  will  be  very  different  according  to  the  average  length 
of  unemployment  spells.  If  the  spells  are  short,  unemployment  is  mostly  transitory.  If  new 
entrants  stay  unemployed  for  a long  period,  their  disadvantage  is  more  acute.  Figure  6 also 
indicates  the  respective  part  of  unemployment  spells  that  last  for  less  and  more  than  one  year. 
It  clearly  shows  that,  apart  from  Greece  and  Italy,  long  spells  are  not  related  to  labour  market 
experience  while  recently  begun  spells  are  experience-dependent.  Greece  and  Italy  differ  as 
they  present  experience-dependent  relationships  for  the  two  types  of  unemployment  spells. 

Mobility  of  new  entrants  on  the  labour  market 

One  can  expect  that  juniors  are  a dynamic  group  on  the  labour  market,  experiencing 
numerous  moves  from  and  to  jobs.  After  leaving  the  ETS,  many  of  them  are  simply  entering 
the  labour  market  without  holding  already  a job.  This  lack  of  established  position  induces  a 
generalised  searching  process.  However,  it  is  worthwhile  to  mention  that  a minority  of  them 
has  already  a job  by  the  time  of  leaving.  They  are  juniors  who  have  got  a job  during  their 
education  and  training  course  and  a part  of  them  keep  this  job  after  leaving  the  ETS.  The 
main  flow  of  those  juniors  staying  within  the  same  company  comes  from  apprenticeship. 
Some  employers  indeed  choose  to  keep  apprentices  they  have  trained  with  a new  labour 
contract4.  However  such  agreements  are  not  a general  rule.  In  addition  to  that  basic  reason, 


4 There  is  no  possibility  of  analysing  such  a flow  in  CLFS. 


others  come  from  the  economic  literature,  applying  on  both  sides  of  the  labour  market,  in 
order  to  explain  the  increased  mobility  of  juniors.  First,  as  presented  in  the  job  search  theory, 
the  process  of  seeking  opportunities  for  a good  job  is  costly.  Lower  level  jobs  may  be  chosen 
in  order  to  reduce  search  costs.  Those  jobs  are  more  or  less  considered  as  short-term  positions 
that  have  to  be  upgraded  later.  Secondly,  even  when  a job  opportunity  appears  to  fulfil 
individual  requirements,  there  is  still  uncertainty  about  the  actual  characteristics  of  the  job. 
Thus  the  quality  of  matching  between  the  individual  and  the  post  has  to  be  established  and 
can  lead  to  job  quit  (Jovanovic  1979).  This  matching  process  is  not  specific  to  the  youngest 
workers  but  one  can  expect  it  to  decline  with  experience.  On  the  other  side  of  the  labour 
market,  the  job  matching  process  also  applies  for  employers,  who  may  not  be  satisfied  with 
newly  hired  people.  The  more  complete  and  comprehensive  the  qualification  signal  is,  the  less 
often  mismatches  occur.  The  less  precise  the  signal  is,  the  more  cautious  employers  will  be. 
Fixed-term  contracts  play  a role  as  a screening  tool  in  order  to  test  performance  on  the  job: 
Apart  from  the  uncertainty  linked  with  the  establishment  of  a new  labour  relationship, 
employers  don’t  always  look  for  permanent  workers.  They  may  prefer  to  use  temporary 
workers  in  the  production  process  in  order  to  regulate  production  activity. 

Although  CLFS  does  not  offer  much  longitudinal  data,  it  is  possible  to  build  indicators  of 
labour  market  changes  between  the  date  of  the  survey  and  one  year  before,  as  is  available  in 
panel  data5.  We  first  study  mobility  between  jobs,  and  then  we  will  focus  on  mobility  between 
employment  and  unemployment. 

Figures  7 & 8 sum  up  two  indicators  of  exits  from  jobs  and  entries  in  jobs.  The  first  is  the 
ratio  of  those  who  had  left  the  job  they  held  last  year  among  all  those  who  held  a job  the  year 
before  while  the  second  is  the  ratio  of  entries  to  jobs  related  to  the  total  group  of  juniors.  As 
the  two  indicators  do  not  have  the  same  basis  of  calculation,  they  are  not  comparable.  The 
entry  measure  reveals  the  relative  ease  with  which  juniors  can  enter  companies,  the  exit 
measure  indicates  the  instability  of  these  job  positions.  The  two  measures  show  among 
European  countries  that  mobility  - both  from  and  into  jobs  - dramatically  and  linearly 


5 Combining  the  current  situation,  the  situation  one  year  before  and  the  potential  seniority  in  job  (computed  from  the  entry 
date  to  the  company),  one  can  construct  different  indicators  of  moves  from  and  to  jobs.  These  indicators  raise  some 
methodological  issues  as  position  on  the  labour  market  is  not  measured  in  the  same  way  at  the  two  dates.  At  the  time  of 
survey,  ILO  criteria  are  applied  to  identify  work  situations,  unemployment  and  inactivity  while  main  situation  is  declared 
by  people  interviewed  for  the  situation  of  the  previous  year.  However,  if  assumptions  that  such  measurement  biases  are 
independent  of  national  context,  age  and  position  on  labour  market  at  each  date  are  verified,  these  indicators  can  be 
interpreted  as  proxies  of  position  shifts.  It  is  useful  to  remember  that  such  indicators  do  not  sum  up  all  of  the  mobility 
that  has  occurred  during  the  past  year  as  numerous  spells  on  the  labour  market  haven’t  been  observed. 


o 

ERIC 


2.28 


200 


declines  as  labour  market  experience  increases.  This  general  trend  can  be  interpreted  as  a 
progressive  stabilisation  process  as  experience  increases. 

Beyond  the  common  tendencies  concerning  job  mobility,  national  peculiarities  still  emerge. 
The  levels  of  moves  in  and  out  vary  strongly  across  countries.  As  the  most  contrasting 
countries,  Spain  is  characteristic  of  massive  junior  mobility  while  Italy  and  Greece  experience 
weak  junior  mobility.  As  a result,  the  Spanish  share  of  job  exits  is  about  three  times  the  share 
of  the  two  others.  Quite  a similar  statement  applies  to  entering  jobs  where  the  Spanish  ratio  is 
twice  the  Greek  one.  However,  the  flexibility  of  junior  activity  appears  to  be  partly  related  to 
flexibility  in  the  national  labour  market.  In  Spain,  the  share  of  job  exits  among  more 
experienced  workers  is  still  three  times  greater  than  the  share  observed  in  Italy  or  Greece.  The 
analysis  of  relative  risks  of  moving  jobs  shows  simultaneously  higher  risks  for  juniors  to  have 
left  jobs  and  to  have  entered  new  jobs.  National  variations  highlight  countries  where  juniors, 
have  a more  vulnerable  situation  in  the  labour  market. 


In  the  same  way,  indicators  of  mobility  between  unemployment  and  jobs  can  be  constructed. 
First,  the  mobility  from  job  to  unemployment,  usually  called  vulnerability  to  unemployment, 
indicates  a relative  fragility  on  the  labour  market  as  it  reveals  the  existence  of  latency  periods 
between  two  jobs.  In  contrast,  mobility  from  unemployment  to  a job  indicates  the  relative 
ease  of  exiting  unemployment.  This  analysis  differs  from  the  previous  analysis  of  mobility  as 
it  is  focuses  on  an  undesirable  event  that  does  not  have  the  same  prevalence  across  EU 
countries  (Figure  9).  Here,  national  situations  fail  to  converge.  Different  configurations 
appear.  The  first  characterises  countries  in  terms  of  weaker  positions  for  juniors  on  the  labour 
market.  The  risk  of  switching  from  employment  to  unemployment  is  increased  for  juniors  in 
Spain,  France,  Sweden  and  United  Kingdom.  The  second  configuration  singles  out 
asymptomatic  conditions  of  switching  between  employment  and  unemployment  for  juniors 
relative  to  their  elders.  Greece  and  Italy  illustrate  this  profile.  The  third  configuration  is 
dominated  by  favourable  prospects  concerning  unemployment  for  juniors.  Compared  to  more 
experienced  workers,  they  have  a greater  ability  to  move  from  unemployment  to  a job, 
without  being  affected  by  a dramatic  increase  in  their  vulnerability.  The  Netherlands 
represents  the  most  typical  example  of  this  profile. 

This  picture  of  mobility  shows  the  LM  operating  with  more  flexible  rules  for  juniors  than  for 
other  categories  of  manpower.  As  a result  of  increased  mobility,  will  it  be  possible  to  relate 


201 


more  flexible  job  positions  to  the  specific  working  conditions  offered  by  companies? 


Work  conditions  of  new  entrants:  more  temporary  contracts  and  involuntary  part-time  jobs 

In  modem  industrial  societies  full-time  salaried  employment  is  dominant.  On  one  hand,  other 
forms  of  employment  -such  as  self-employment  and  family  worker  status-  have  long  been 
relegated  to  specific  economic  activities  or  occupations.  On  the  other  hand,  part-time  work, 
albeit  expanding,  still  affects  a minority  of  workers  even  if  its  relative  importance  varies 
among  EU  countries.  Again,  among  different  forms  of  labour  contracts,  the  full-time 
permanent  one  has  emerged  as  the  standard  type  of  employee/employer  relationship. 
Although  the  labour  legislation  concerning  these  full-time  'permanent'  contracts  differs  from 
one  EU  country  to  another  in  various  ways  (the  trial  period  specified,  working  hours 
legislation  or  redundancy  rules,  among  other  elements),  it  always  offers  a national  standard 
platform  for  the  labour  relationship.  Opposite  to  this  standard,  several  particular  forms  of 
salaried  jobs  have  arisen.  Among  them,  the  two  most  common  forms  are  fixed-term  (or 
'temporary')  contracts  and  part-time  activity.  Still,  their  prevalence  still  depends  on  the 
restrictive  nature  of  national  regulations. 

We  focus  here  on  the  relative  importance  of  those  two  forms  of  salaried  employment.  The 
involuntary  nature  of  these  employment  conditions  will  be  stressed  as  an  indication  of  new 
entrants’  dissatisfaction  with  their  employment  outcomes. 

Temporary  contracts  as  a specific  mode  of  hiring  new  entrants 

Temporary  contracts  are  not  equally  developed  among  EU  countries.  The  case  of  Spain, 
where  they  have  a massive  impact,  differs  strongly  from  other  countries,  as  Austria  or 
Luxembourg,  where  they  hardly  exist.  Countries  where  these  employment  contracts  are 
widespread,  have  a common  profile  (Figure  10).  It  is  characterised  by  high  initial  levels  for 
the  least  experienced  new  entrants  with  a linear  decline  as  experience  increases.  The  Spanish 
case  has  to  be  singled  out,  as  this  form  of  contract  represents  the  mainstream  labour 
relationship  for  juniors.  It  only  falls  below  50%  of  salaried  contracts  after  about  seven  years 
of  LM  experience.  Involuntary  temporary  contracts  are  more  prevalent  in  countries  where 
temporary  contracts  are  commonly  used. 


2.30 


202 


Figure  7:  Ratio  of  job  exits  (people  having  left  their  job  among  people  with  same  LM  experience  who  held  a job 

one  year  before) 


Figure  9:  mobility  between  unemployment  and  employment,  by  experience  on  the 

labour  market 


Figure  9a  : “OLM-type”  countries 


Figure  9b:  Southern  countries 


E L 

1 " 

i r 
* 1 ' 

i . 

• • 

► 

% 



-i  lKKi*-“n 

IT 

$ = 

2.32 


205 


Figure  9c:  Other  countries 


2.33 


206 


Figure  10:  Temporary  contracts  and  years  of  experience  on  the  labour  market  among 

employees 


Figure  10a  : “OLM-type”  countries 


80 

70 


D E 35  - 97 


60 


50 

40 

30 

20 

10 

0 


t 2 1 4 S I 7 I J 1011  12  1)141$ 


Figure  10b:  Southern  countries 


y 0 ther  jo  bs  with  temporarycontracts  g Jo  bs  with  unwilling  te  m po  rary  co  ntrat 


Jobs  with  all  kinds  o f tem  po  rary  co  n tracts 


2.34 


20,7 


Figure  10c:  Other  countries 


8 0 

7 0 
60 

UK  95  ■ 97 

5 0 

40 
30  - 

20  - 

10 

1 2 1 4 5 • 7 t • 1911  121)141 

5 

ggOthor  jo  bs  with  tem  po  rary  co  n tracts  ■ Jo  bs  with  unwilling  tem  po  rary  co  ntrat  Jobs  with  all  kinds  o f tem  po  rary  co  n tracts 


208 


Figures  11:  Part-time  contracts  and  years  of  experience  on  the  labour  market  among 

employees 


Figure  11a:  “OLM-type”  countries 


Figure  lib:  Southern  countries 


13  Other  jobs  with  part-time  contracts 


■ Jobs  with  unwilling  part-time  contrat 


Figure  11c:  Other  countries 


□ Other  jobs  with  part-time  contracts  aJobs  with  unwilling  part-time  contrat 


O 

ERIC 


2.37 


210 


Part-time  jobs  and  new  entrants:  a link  structured  by  the  involuntary  dimension  of  the  labour 
relationship 


As  for  temporary  contracts,  the  extent  of  part-time  jobs  is  fairly  dependent  on  national 
context.  Commonly  used  in  northern  countries  (Denmark,  the  Netherlands  and  Sweden),  they 
are  more  unusual  in  southern  countries  (Portugal,  Greece  and  Spain).  However,  their  impact 
on  new  entrants’  situation  differs  from  that  of  temporary  contracts.  The  trend  is  rather  the 
reverse;  the  level  of  part-time  jobs  remains  at  least  constant  as  experience  grows,  and  in  fact 
generally  increases  with  the  accumulation  of  experience  (Figure  11).  In  some  countries, 
where  they  are  extensive,  their  frequency  according  to  experience  shows  a U-shaped  curve 
with  first  a slight  decrease  and  then  a recovery.  As  a result,  part-time  jobs  as  a whole  do  not 
present  a specific-dependent  link  with  labour  market  outcomes  for  new  entrants. 

Introducing  the  subjective  involuntary  dimension  of  such  a relationship  leads  us  to  modify  our 
results.  These  kind  of  part-time  jobs  vary  in  their  prevalence  across  countries.  Almost  absent 
in  Luxembourg  and  Austria,  they  reach  their  maximum  level  in  France,  Sweden  and  Finland. 
In  countries  with  a significant  initial  level  of  involuntary  part-time  work,  its  level  declines 
with  experience.  Moreover,  as  experience  increases,  the  decrease  in  involuntary  part-time 
jobs  is  balanced  by  the  increase  in  other  kinds  of  part-time  jobs. 

Quality  and  characteristics  of  jobs:  which  occupations  and  companies  for  new 
entrants? 

Apart  from  the  specific  features  of  employment  access  and  employment  conditions, 
companies  allocate  new  entrants  to  specific  activities.  First,  companies  that  choose  to  hire 
new  entrants  have  specific  features.  Second,  the  positions  held  by  new  entrants  within 
companies  are  somewhat  distinctive.  The  way  employers  reward  the  qualifications  achieved 
by  new  entrants  is  a third  issue. 


ERIC 


2.38 


Companies:  private  individual  services  and  business  services  are  the  most  favourable 
economic  activities  for  new  entrants 

The  main  issue  about  companies  hiring  new  entrants  is  to  determine  whether  or  not  they  have 
particular  recruitment  policies.  If  it  is  so,  companies  open  to  new  inexperienced  workers  may 
be  differentiated  from  the  others.  As  a result,  new  entrants  should  be  concentrated  in  a 
specific  type  of  companies.  However,  it  is  important  to  keep  in  mind  that  the  available 
information  (CLFS)  is  basically  stock  data.  It  means  that  shifts  in  the  national  structure  of 
companies  cannot  be  taken  into  account. 

First,  the  industrial  allocation  of  juniors  shows  some  convergent  features  across  EU  countries 
(table  3).  Tertiary  activities  appear  to  be  dominant  in  all  countries  while  agriculture  has  a 
reduced  impact  in  most  of  them.  Among  tertiary  activities,  wholesale  and  retail  trade  has  a 
strong  impact.  Nevertheless,  national  peculiarities  are  numerous.  The  most  striking  examples' 
can  be  summarised  briefly.  Agriculture  is  important  in  Greece  while  industry  is  most 
developed  among  working  juniors  in  Italy,  Portugal  and  Ireland.  Financial  services  are  well 
developed  in  Luxembourg.  Health  services  have  an  important  weight  in  Sweden,  Denmark, 
Finland  and  the  Netherlands.  These  examples  clearly  show  that  the  economic  activity  of 
companies  hiring  juniors  partly  reflects  the  national  structure  of  economic  activities. 

Table  3:  Structure  of  economic  activities  for  juniors  - average  95-97 


% 


Economic  activities  (Nace) 

DK 

DE 

NL 

AT 

EL 

ES 

IT 

PT 

BE 

FR 

IE 

LU 

FI 

SE 

UK 

EU 

15 

Agriculture 

3 

2 

4 

3 

12 

5 

7 

5 

2 

3 

5 

3 

4 

2 

2 

4 

Industry  (excluding  construction) 

21 

22 

15 

23 

16 

21 

36 

30 

22 

21 

27 

10 

24 

22 

21 

23 

Construction 

5 

9 

7 

9 

6 

10 

9 

13 

fi7 

5 

7 

9 

4 

5 

6 

8 

Wholesale  & retail  trade 

17 

14 

22 

20 

24 

19 

18 

15 

16 

17 

17 

17 

14 

17 

19 

18 

Hotels  & Restaurants 

3 

4 

4 

8 

9 

7 

6 

6 

5 

5 

8 

5 

4 

7 

7 

5 

Transport  & communication 

6 

4 

5 

5 

4 

3 

2 

3 

6 

4 

2 

5 

5 

6 

5 

4 

Financial  services 

3 

5 

4 

3 

3 

3 

2 

2 

4 

3 

4 

14 

1 

2 

6 

4 

Business  activities 

9 

8 

12 

6 

7 

9 

7 

7 

9 

13 

8 

8 

8 

11 

10 

9 

Public  administration 

6 

10 

6 

5 

5 

4 

3 

4 

6 

6 

r 3 

11 

4 

3 

r 4 

6 

Education 

6 

5 

4 

5 

7 

6 

2 

7 

10 

7 

5 

6 

10 

5 

^ 4 

5 

Health 

16 

10 

13 

7 

3 

5 

3 

4 

11 

9 

7 

8 

15 

17 

8 

[r  8 

Other  service  activities 

4 

5 

4 

5 

6 

7 

5 

5 

4 

6 

6 

5 

6 

4 

6 

| 6 

All 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

Source:  Eurostat,  LFS 


Indeed,  once  the  national  structure  is  taken  into  account,  the  pattern  shifts  somewhat.  Table  4 
shows  the  relative  concentration  of  juniors  across  different  economic  activities.  The  index  is 
calculated  by  comparing  the  share  of  juniors  in  a specific  economic  activity  with  their  share  in 
total  employment.  The  results  strengthen  our  initial  impression  of  convergence  among 
countries  although  they  indicate  that  different  economic  activities  play  an  important  role  in 
juniors’  employment.  By  and  large,  private  individual  and  business  services  appear  to  have  a 
dominant  role.  Hotels  and  restaurants  have  the  highest  degree  of  over-representation  of 
juniors,  followed  by  business  activities,  wholesale  & retail  trade  and  financial  services.  On 
the  other  hand,  agriculture,  transport  & communication,  public  administration  and  Education 
are  under-represented  activities  in  juniors’  employment.  Industry,  construction  and  health 
show  a balanced  representation  of  juniors.  Some  converging  trends  among  EU  countries 
emerge:  under-representation  of  juniors  in  agriculture  for  almost  all  the  countries  and  general 
over-representation  in  hotels  & restaurants.  Industry,  construction,  education  and  financial 
services  display  more  diversity  across  Europe.  Further  analysis  taking  account  of  internal 
business  cycle  of  economic  activities  would  partly  help  to  explain  such  variations.  For 
example,  the  period  of  1995  to  1997  was  one  of  a booming  business  cycle  in  Ireland, 
especially  in  manufacturing  industry. 


Table  4:  relative  concentration  of  juniors  in  economic  activities  - average  1995-1997 


Economic  Activities 

DK 

DE 

NL 

AT 

EL 

ES 

IT 

PT 

BE 

FR 

IE 

LU 

FI 

SE 

UK 

EU 

15 

Ratio 

EU-15 

Agriculture 

- 

- 

= 

— 

- 

- 

= 

— 

- 

- 

... 

- 

- 

- 

- 

0,69 

Industry  (excluding 
construction) 

= 

- 

- 

= 

= 

— 

++ 

++ 

= 

++ 

- 

= 

+ 

= 

= 

1,06 

Construction 

- 

= 

= 

+ 

= 

= 

= 

++ 

= 

- 

- 

= 

- 

- 

- 

= 

0,98 

Wholesale  & retail  trade 

+ 

= 

-H- 

+ 

-H- 

+ 

= 

= 

= 

+ 

+ 

+ 

+ 

++ 

+ 

+ 

1,16 

Hotels  & Restaurants 

+ 

+ 

+ 

++ 

-H-+ 

+ 

-H- 

+ 

++ 

-H-+ 

++ 

+ 

+++ 

+++ 

++ 

++ 

1,35 

Transport  & communication 

- 

- 

- 

- 

- 

- 

— 

- 

- 

- 

- 

- 

- 

- 

- 

- 

0,73 

Financial  services 

= 

++ 

+ 

- 

- 

- 

- 

+ 

++ 

- 

- 

++ 

+ 

1,12 

Business  activities 

+ 

+ 

+ 

= 1 

+++ 

-H-+ 

+ 

-H-+ 

++ 

++ 

++ 

++ 

= 

+ 

+ 

1,26 

Public  administration 

= 

+ 

- 

- 

- 

- 

— 

- 

- 

- 

— 

- 

- 

— 

- 

- 

0,77 

Education 

- 

= 

— 

- 

+ 

+ 

— 

= 

= 

- 

= 

++ 

- 

— 

- 

0,78 

Health 

= 

+ 

- " 

- 

- 

- 

- 

- 

0,91 

Other  service  activities 

- 

= 

- 

+ 

+ 

= 

= 

- 

= 

= 

= 

= 

+ 

- 

+ 

= 

1,00 

Source:  Eurostat,  LFS 


Reading  guide  : 

= if  the  ratio  between  the  share  of  juniors  in  the  economic  activity  and  the  share  of  juniors  in  all  the  economic  activities  is 
between  0,9  and  1,1. 

+ if  the  ratio  is  between  1,1  and  1,3. 

++  if  the  ratio  is  between  1,3  and  1,5. 

+++  if  the  ratio  is  over  1,5. 

- if  the  ratio  is  between  0,7  and  0,9. 

- if  the  ratio  is  between  0,5  and  0,7. 

- if  the  ratio  is  under  0,5. 

positive  signs  indicate  an  over-representation  of  juniors  in  economic  activities,  negative  signs  indicate  an  under- 
representation of  juniors. 


Beyond  the  focus  of  juniors’  employment  towards  specific  economic  activities,  the  issue  of  an 
overall  trend  of  concentration  in  specific  businesses  remains.  Table  5 analyses  an  indicator  of 
such  a measure  of  concentration.  It  presents  the  coefficient  of  variation  of  juniors’  share 
among  economic  activities.  This  coefficient  equals  the  ratio  of  the  standard  deviation  of 
juniors’  share  by  its  mean.  The  greater  the  coefficient,  the  wider  is  the  dispersion  of  juniors’ 
share  across  economic  activities.  A high  value  indicates  a tendency  of  juniors  to  be 
concentrated  in  some  economic  activities  while  a low  value  conversely  indicates  that  juniors 
are  more  widely  spread  out  across  economic  activities.  The  results  indicate  significant 
variations  between  countries.  On  one  side,  Italy  and  Greece  show  a high  degree  of 
concentration  relative  to  other  countries.  On  the  other  side,  Germany  and  Denmark  show  a 
low  degree  of  concentration.  This  can  be  interpreted  as  a more  open  economy  in  the  latter 
case  and  a more  stratified  economy  with  new  entrants  confined  to  certain  economic  activities 
in  the  former. 


Table  5:  Coefficient  of  variation  of  juniors'  share  across  economic  activities  (all  %) 


Standard  deviation 

mean  of  juniors  share 
in  the  economic 
activity 

Coefficient  of 
variation  - 1997 

DK 

2,0 

14,0 

17,7 

DE 

2,0 

9,0 

18,6 

NL 

2,0 

9,0 

21,6 

AT 

2,0 

10,0 

23,8 

EL 

3,0 

9,0 

37,9 

ES 

3,0 

11,0 

27,5 

IT 

4,0 

9,0 

42,6 

PT 

3,0 

8,0 

38,8 

BE 

2,0 

10,0 

23,6 

FR 

2,0 

8,0 

28,0 

IE 

5,0 

15,0 

35,5 

LU 

3,0 

9,0 

32,9 

FI 

2,0 

9,0 

25,3 

SE 

3,0 

8,0 

38,6 

UK 

2,0 

9,0 

28,1 

Source:  Eurostat,  LFS 


Having  taken  a look  at  the  profile  of  companies  hiring  them,  the  focus  is  now  on  the 
occupations  new  entrants  hold. 


uc 


2.41 


214 


Occupations:  the  importance  of  service  workers  and  sales  workers  for  new  entrants 


The  main  issue  raised  by  the  occupational  structure  of  new  entrants'  jobs  relates  to  potential 
occupations.  Does  the  integration  process  imply  open  or  closed  occupations  to  new  entrants? 
The  national  occupational  structure  of  employed  juniors,  using  the  ISCO  classification, 
appears  to  be  largely  influenced  by  economic  activities  (table  6).  Although  there  is  no  strict 
correspondence  within  companies  between  their  economic  activity  and  their  occupational 
positions,  the  links  are  quite  strong.  Two  examples  show  the  existence  and  the  limits  of  such 
links.  In  Greece,  agricultural  activity  represents  12%  of  juniors’  employment,  while  skilled 
agricultural  workers  are  at  the  same  level.  We  have  earlier  pointed  out  that  industry  is  the 
most  developed  in  Italy,  Portugal  and  Ireland  among  working  juniors.  We  now  observe  that 
blue  collar  groups  (ISCO  7 to  9)  are  the  most  developed  in  Italy  and  Portugal,  followed  by 
Ireland,  Spain,  Austria  and  Sweden.  The  three  latter  countries  nevertheless  have  a less 
significant  share  of  manufacturing  industry  activity.  Nevertheless,  national  occupational 
structures  show  wide  variations  from  one  country  to  another.  The  relative  importance  of 
upper  white  collar  groups  (technicians,  professionals  and  managers,  ISCO  1 to  3)  varies 
dramatically  between  Italy  (16%  of  juniors'  jobs)  and  Finland  (43%).  However,  these 
variations  again  partly  reflect  the  differentiation  of  the  overall  occupational  structure. 


Table  6:  Structure  of  occupations  for  Juniors  - average  1995-1997 


% 


Occupations  (ISCO-88,  first  digit) 

DK 

DE 

NL 

AT 

EL 

ES 

IT 

PT 

BE 

FR 

IE 

LU 

FI 

SE 

UK 

EU 

15 

l=Managers 

4 

2 

r 4 

3 

4 

2 

0 

2 

6 

5 

4 

2 

5 

1 

8 

4 

2=Professionals 

16 

17 

15 

9 

13 

17 

5 

13 

25 

14 

17 

20 

26 

14 

11 

14 

3=Technicians  & associated 
professionals 

17 

21 

19 

12 

7 

7 

11 

9 

10 

19 

4 

16 

12 

14 

9 

14 

4=Clerks 

14 

14 

15 

16 

14 

12 

15 

10 

16 

16 

17 

20 

5 

11 

23 

16 

5=Service  workers  & sales  workers 

18 

14 

17 

20 

24 

19 

18 

16 

13 

15 

22 

15 

16 

23 

19 

17 

6=skilled  agricultural  workers 

2 

1 

2 

2 

12 

2 

2 

2 

2 

3 

2 

3 

3 

1 

1 

2 

7=Craft  & related  trades  workers 

12 

20 

12 

25 

16 

14 

28 

30 

14 

11 

17 

15 

16 

13 

12 

17 

8=Plant  & machine  operators 
& assemblers 

7 

5 

7 

7 

5 

8 

11 

5 

7 

9 

8 

4 

10 

14 

7 

7 

9=Elementary  occupations 

10 

6 

8 

6 

5 

17 

8 

13 

7 

7 

9 

7 

7 

10 

10 

9 

All 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

100 

Source:  Eurostat,  LFS 


If  the  relative  concentration  is  examined  (Table  7),  they  are  found  to  be  concentrated  in  two 
occupational  groups  - service  workers  and  sales  workers.  Clerks  and  professionals  represent 
two  groups  open  to  juniors,  except  in  some  countries  (Italy  for  instance).  Conversely,  juniors 


O 

ERIC 


2.42 


21 5 


are  strongly  under-represented  among  skilled  agricultural  workers  and  managers  and,  to  a 
lesser  extent,  the  machine  operators  and  assemblers  group.  Other  groups  (technicians,  craft 
workers  and  elementary  occupations)  have  a more  balanced  representation  of  juniors  at  the 
European  level.  However,  national  comparisons  reveal  a high  degree  of  differentiation 
between  countries  in  these  groups.  The  national  summaries  show  impressive  contrasts 
between  countries.  Besides  Denmark  showing  the  least  marked  variation  of  juniors  in  the 
occupational  structure,  Italy  and  France  present  two  contrasting  patterns.  Juniors  are  over- 
represented in  blue-collar  groups  (apart  from  elementary  occupations)  in  Italy  but  under- 
represented among  these  groups  in  France.  Conversely,  French  upper  white  collar  positions 
(apart  from  managers)  represent  an  important  outlet  for  juniors  while  such  positions  are  more 
or  less  closed  to  them  in  Italy. 


Table  7:  Relative  concentration  of  juniors  in  occupations  - average  1995-1997 


Occupations  (ISCO  88) 

DK 

DE 

NL 

AT 

EL 

ES 

IT 

PT 

BE 

FR 

IE 

LU 

FI 

SE 

UK 

15 

EU-15 

l=Managers 

- 

— 

— 

--- 

— 

— 

— 

— 

- 

- 

— 

— 

— 

— 

— 

— 

0,46 

2=Professionals 

+ 

++ 

= 

- 

= 

+++ 

— 

+++ 

++ 

++ 

= 

++ 

++ 

= 

- 

+ 

ui 

3=Technicians  & 
associated  professionals 

= 

= 

= 

= 

+ 

= 

- 

- 

— 

+ 

— 

— 

- 

- 

= 

= 

0,99 

4=Clerks 

= 

= 

+ 

+ 

+ 

+ 

= 

= 

= 

+ 

+ 

+ 

— 

= 

++ 

+ 

1,14 

5=Service  workers 
& sales  workers 

+ 

+ 

++ 

+++ 

+++ 

++ 

+ 

+ 

++ 

+ 

+ 

++ 

+ 

++ 

++ 

+ 

1,28 

6=skilled  agricultural  workers 

- 

- 

++ 

= 

— 

— 

- 

- 

0,55 

7=Craft  & related  trades 
workers 

= 

= 

+ 

++ 

= 

- 

++ 

++ 

= 

- 

++ 

= 

+ 

= 

= 

= 

1,06 

8=Plant  & machine  operators 
& assemblers 

- 

— 

= 

- 

— 

- 

+ 

- 

- 

- 

- 

— 

+ 

+ 

- 

- 

0,83 

9=Elementary  occupations 

- 

- 

= 

- 

- 

+ 

- 

+ 

- 

= 

= " 

- 

- 

+++ 

+ 

= 

0,93 

Source:  Eurostat,  LFS 

Reading  guide  : 

= if  the  ratio  between  the  share  of  juniors  in  the  occupational  group  and  the  share  of  juniors  in  all  the  occupations  is  between 
0,9  and  1,1. 

+ if  the  ratio  is  between  1,1  and  1,3. 

-H-  if  the  ratio  is  between  1,3  and  1,5. 

+++  if  the  ratio  is  over  1,5. 

- if  the  ratio  is  between  0,7  and  0,9. 

- if  the  ratio  is  between  0,5  and  0,7. 

- if  the  ratio  is  under  0,5. 

positive  signs  indicate  an  over-representation  of  juniors  in  occupational  groups,  negative  signs  indicate  an  under- 
representation of  juniors. 


216 


2.43 


More  about  the  Links  between  the  ETS  and  the  LM:  some 
insights  on  the  importance  of  educational  attainment  in  the 
transition  from  school-to-work 

Until  now,  we  have  focused  on  the  general  patterns  of  labour  market  entry  in  the  EU 
countries,  restricting  the  contribution  of  educational  background  to  the  issue  of  the 
biographical  timing  of  leaving  ETS.  This  has  led  us  to  neglect  the  importance  of  educational 
attainment.  Here  we  want  to  briefly  outline  some  individual  effects  of  educational  attainment 
in  the  school-to-work  transition.  We  focus  only  on  a few  indicators  to  stress  the  key  role  of 
education  in  LM  outcomes.  Firstly,  we  study  the  impact  of  educational  attainment  on 
unemployment  risks.  Secondly,  the  mobility  between  jobs  and  unemployment  is  presented. 
Thirdly,  we  concentrate  on  the  returns  to  educational  qualifications  across  occupations. 

Unemployment  by  level  of  education  attained 

Introducing  the  educational  attainment  of  new  entrants  in  the  analysis  of  the  unemployment 
risk  by  labour  market  experience  yields  a more  scattered  picture  than  is  evident  at  a more 
general  level  (Figure  12).  Three  configurations  emerge  in  relation  to  the  impact  of  the  two 
dimensions.  The  first  one  is  mainly  ruled  by  the  dominant  effect  of  educational  level 
attainment  and  a corresponding  residual  effect  of  experience  on  the  unemployment  rate. 
Conversely,  the  second  is  mainly  characterised  by  the  influential  role  of  LM  experience  in 
conjunction  with  a weakened  qualification  effect  while  the  third  is  influenced  by  both 
dimensions.  Austria,  the  Netherlands,  Denmark  and  somewhat  Germany  belong  to  the  first 
configuration.  They  show  stratified  levels  of  unemployment  by  educational  level  attained, 
fairly  independently  of  accumulated  experience.  Italy  and  Greece  are  the  most  representative 
countries  of  the  second  group.  They  show  declining  levels  of  unemployment  as  experience 
increases  with  a rather  blurred  distinction  between  ISCED  levels.  Other  countries  are  more  or 
less  close  to  the  third  configuration  combining  the  two  effects.  Belgium,  France,  Ireland, 
Sweden  and  United  Kingdom  are  the  most  typical  countries  within  this  group.  They  show 
declining,  well-differentiated  unemployment  rates  by  ISCED  level  as  experience  increases. 


ERIC 


Figure  12:  Observed  Unemployment  Rate  and  Adjusted  Unemployment  Rate  according 

to  years  of  LM  experience  and  ISCED  level 

(U=l/(l+exp(a+b,T+b2E+b3E2+b4TE+b5TE2))  with  T=  Training  level  and  E=LM  experience) 


Figure  12a:  “OLM-type”  countries 


Figure  12b:  Southern  countries 


70%  - 

EL  93  - 97 

60%  - 

50%  J 

o 

40% 

* 

30% 

to  j.  * ♦ 

vi  ® * 

20% 

10% 

i r i ■ 

0%  . 

■ Observed  U rate  (Isced  0-2)  A djusted  U rate  (Isced  0-2)  © 0 bserved  U rate  (Isced  3) 

Adjusted  U rate  (Isced  3)  A 0 bserved  U rate  (Isced  5-7)  A djusted  U rate  (Isced  5-7 ) 


2.45 


218 


Figure  12c:  Other  countries 


■ Observed  U rate  (Isced  0-2) 
. ...  A djusted  U rate  (isced  3) 


A djusted  U rate  (Isced  0*2)  o Observed  U rate  (Isced  3) 

A Observed  U rate  (Isced  5-7)  . -A  djusted  U rate  (Isced  5-7) 


i*- 


2.46 


Figure  13:  Transitions  from  employment  to  unemployment  and  from  unemployment  to 

employment  for  Juniors,  by  ISCED 


C 

Q> 

E 

>» 

o 

Q. 

E 

a> 

c 

3 

o 

* 


A 

ns 

u 

a> 

c 


> 


% having  a job  among  people  unemployed  1 year  ago 


♦ ISCED  5-7  ■ ISCED  3 a ISCED  0-2  ^European  Union 


o 

ERLC 


2.47 


220 


Mobility  between  unemployment  and  employment  by  level  of  education  attained 

Going  back  to  the  analysis  of  the  mobility  between  unemployment  and  employment, 
vulnerability  to  unemployment  declines,  and  employment  prospects  for  the  unemployed 
increase  with  the  level  of  qualification  (Figure  13).  It  confirms  the  hierarchy  of 
unemployment  risks  previously  established.  Moreover,  this  outcome  is  evident  for  most  of  the 
countries  at  the  highest  level  of  qualification.  Only  two  countries  show  very  different  profiles. 
Italy  and  Greece  are  characterised  by  very  little  differentiation  between  education  levels  in 
risks  and  prospects  with  relative  independence  between  these  LM  outcomes  and  qualification 
attainment  inside  the  ETS. 


Occupational  returns  to  qualifications 


The  returns  to  qualifications  achieved  inside  the  ETS  are  to  be  questioned.  It  deals  with  the 
way  the  grading  of  skills  and  qualification  structures  interact. 

Establishing  returns  to  qualifications  necessitates  having  a measure  of  occupation  payoff. 
Wages  are  a natural  measure  of  such  a payoff,  but  they  do  not  represent  the  only  relevant 
dimension  of  the  rating  of  an  occupation.  Besides  work  conditions  (including  job  stability  and 
possible  hourly  constraints  among  others)  occupations  are  ranged  on  a social  scale.  Different 
ways  have  been  developed  in  order  to  devise  a graded  structure  of  jobs  based  either  on 
subjective  valuation  such  as  the  prestige  associated  with  the  job  (SIOP  index)  or  on  more 
objective  criteria  such  as  the  socio-economic  structure  of  the  occupational  group 
(International  Socio-Economic  Index  or  ISEI,  see  Ganzeboom  and  Treimann  (1996)  for  the 
conversion  of  ISCO  into  ISEI).  Given  the  availability  of  wages  information  for  juniors  in  a 
limited  set  of  EU  countries  (extracted  from  the  Wages  Structure  Survey,  1995  from  Eurostat), 
we  have  therefore  two  somewhat  complementary  measures  of  job  payoff  (Figure  14).  The 
figure  represents,  for  the  two  measures,  the  index  calculated  by  relating  the  mean  of  the  ISEI 
score  (respectively  wages)  observed  among  a junior  group  of  a given  qualification  level  to  the 
national  mean  observed  among  all  workers.  It  shows  that,  in  terms  of  the  ISEI  index,  there  is 
a sharp  divide  between  juniors  with  higher  education  levels  (ISCED  5-7)  and  other  juniors. 
In  some  cases  (Denmark,  Germany,  Austria,  Ireland,  Italy,  Greece,  Spain  and  Portugal) 
another  less  marked  divide  can  be  observed  between  juniors  having  achieved  upper  secondary 


2.48 


221 


education  or  training  (ISCED  3)  and  the  others  (ISCED  0-2).  However,  the  wages  analysis  on 
available  countries  only  partially  confirms  such  results.  Higher  education  is  still  associated 
with  an  improved  payoff,  but  the  gap  is  more  or  less  reduced  in  all  countries.  Moreover,  the 
high  reward  - above  the  national  mean-  measured  by  the  ISEI  score  in  Spain,  Ireland,  Italy 
and  Greece  is  contradicted  by  the  wages  index.  This  leads  us  to  conclude  that  even  the 
occupational  groups  entered  by  higher  educated  juniors  are  highly  ranked  and  differ  from 
those  entered  by  juniors  from  upper  secondary  education.  As  a consequence,  their  monetary 
payoff  tends  to  be  sharply  reduced  relative  both  to  upper  secondary  juniors  and  to  more 
experienced  workers.  On  the  opposite  side,  the  Danish  and  Austrian  monetary  returns 
structure  confirms  the  status  (ISEI)  returns  structure  being  graded  by  ISCED  level. 


Figure  14:  Index  of  Juniors'  average  wages  and  ISEI  scores  by  level  of 
education  (100=  national  means  among  all  workers) 


Figure  14a  : “OLM-type”  countries 


Figure  14b:  Southern  countries 


2.50 


P?Q 


Figure  14c:  Other  countries 


LU 

CO 


3 


224 


Figure  15:  Average  ISEI  score  by  ISCED  level  and  years  of  experience, 

average  1995-1997 


Figure  15a:  “OLM-type”  countries 


Figure  15b:  Southern  countries 


7 5 

EL  93  - 97 

6 S 

5 5 

4 5 

35 
2 5 

\ : 

v. 

1 2 1 4 S 1 7 | « 10  11  12  1 1 11  IS 

75  - 

IT  93  - 97 

6 5 

— . - — 

5 5 

/ 

4 5 

35 

25  - 

1 2 1 4 J 1 2 1 1 1011  12111415 

2.52 


225 


Figure  15c:  Other  countries 


1 t 1 4 J I 7 | | 1011  171)141$ 


ISCED  0-2  ISCED  3 ISCED  5-7  — - . All  ISCED 


226 


The  cross-analysis  of  qualification  returns  by  accumulated  years  of  LM  experience  shows 
again  the  clear  differentiation  of  qualifications  returns  according  to  the  ISEI  score  (Figure 
15).  A common  feature  across  all  EU  countries  is  the  flatness  of  the  ISCED  0-2  ISEI  curve 
associated  with  very  low  levels  of  ISEI.  This  indicates  the  lack  of  possible  promotion  to  more 
skilled  occupations.  However,  distinctive  profiles  can  be  identified  on  the  basis  of  ISCED  3 to 
7 new  entrants.  First,  several  countries  have  flat  slopes,  an  indication  of  the  independence 
between  ISEI  grading  and  the  process  of  experience  accumulation  among  the  qualification 
groups.  Germany,  Denmark,  Austria  and  Greece  exhibit  this  profile.  Denmark  and  Germany 
are  even  atypical  as  a declining  trend  appears  with  the  accumulation  of  experience  for  new 
entrants  of  ISCED  5-7  level.  A fairly  simple  explanation  can  be  proposed  for  that  situation  ; 
as  soon  as  juniors  of  ISCED  3 level  have  accumulated  a minimum  number  of  years  of  labour 
experience,  they  obtain  the  opportunity  to  prepare  a Tecknischer  or  Meister  certificate  and  to 
attain  a ISCED  5 level.  Then  they  enter  a new  segment  of  the  labour  market  to  occupy 
technician  positions  with  the  lowest  ISEI  values  among  the  upper  white  collar  group. 
Meanwhile,  juniors  having  left  the  initial  higher  education  system  with  ISCED  6-7  have 
already  entered  the  labour  market  where  they  have  got  occupations  with  a high  ISEI  score. 
The  conjunction  of  the  two  groups  results  in  a declining  effect  on  the  ISEI  score.  Other 
countries  show  profiles  with  a gradual  upgrading  of  ISEI  score  for  one  or  both  new  entrants 
groups.  The  United  Kingdom  and  Ireland  are  clear  examples  of  ISCED  5-7  upgrading,  Spain 
represents  the  case  of  ISCED  3 upgrading  while  Italy  shows  a simultaneous,  quite  strong,  rise 
of  ISEI  score  for  the  two  groups.  In  every  case,  there  is  an  interaction  between  qualification 
returns  and  accumulated  experience. 


Conclusion 


Here  we  have  roughly  outlined  some  characteristics  of  labour  market  outcomes  that  occur 
during  the  school-to-work  transition.  First,  the  specificity  of  transitory  events,  such  as 
combined  training-and-working  situations  or  high  unemployment,  has  been  assessed.  Second, 
the  switch  from  youngsters  to  new  entrants  has  highlighted  the  variable  role  of  experience  in 
labour  market  outcomes.  Third,  the  differentiation  of  rewards  by  qualifications  attainment 


ERIC 


2.54 


22  7 


inside  the  ETS  underlines  the  existence  of  strong  links  between  the  ETS  and  the  economic 
system.  Fourth,  LM  organisation  proves  to  have  an  impact  on  the  shape  of  LM  outcomes  for 
new  entrants  as  shown  by  the  differentiated  role  of  legislation  concerning  work  contracts.  It 
raises  some  issues  about  the  nature  of  the  school-to-work  transition. 

The  respective  roles  of  experience  and  qualification  raise  the  issue  of  interaction  between  new 
entrants’  allocation  on  the  labour  market  and  the  general  organisation  of  employment 
relationships.  Relations  and  competition  between  new  entrants  and  more  experienced  workers 
indicated  both  common  trends  across  Europe  and  specific  national  features  linked  to 
institutional  arrangements.  The  variation  of  outcomes  among  the  new  entrants  group  led  to 
similar  conclusions. 

Features  of  the  national  economic  context  have  not  been  taken  into  account.  As  this  paper  has 
been  mainly  based  on  cross-sectional  data  (with  a limited  set  of  longitudinal  information) 
summarised  over  a three-year  span,  it  cannot  describe  the  dynamics  of  the  process.  On  the 
labour  demand  side,  the  state  of  the  business  cycle  or  the  structural  shift  in  labour  demand 
have  not  been  analysed.  In  the  same  way,  structural  changes  in  the  labour  supply  such  as 
demographic  pressure  or  qualification  shifts  have  not  been  dealt  with. 

The  issue  of  the  impact  of  national  public  policies  on  labour  market  has  not  been  dealt  with. 
There  have  been  recently  important  changes  in  some  countries.  Unfortunately,  CLFS  prior  to 
1998  have  too  limited  a set  of  information  about  such  policy  changes.  Improvements  in  more 
recent  data,  including  the  2000  transition  module,  will  yield  material  for  further  analysis. 


Bibliography 

Ganzeboom  H.,  and  D.  Treiman,  1996:  “Internationally  Comparable  Measures  of 
Occupational  Status  for  the  1988  International  Standard  Classification  of  Occupations”. 
Social  Science  Research  25, 201-239. 


Hannan  D.,  D.  Raffe  and  E.  Smyth,  1997  : “Cross-national  research  on  School  to  Work 
Transitions:  an  analytic  framework”,  in  Youth  Transitions  in  Europe:  Theories  and  Evidence, 
P.  Werquin,  R.  Breen  and  J.  Planas  (eds.),  Cereq,  Marseilles. 

0 


ERLC 


l 


2.55 


228 


Hannan  D.  and  P.  Werquin  1999  : “Education  and  labour  market  change:  the  dynamics  of 
education  to  work  transitions  in  Europe”,  European  Socio-Economic  Conference,  Brussels, 
April. 

Hannan,  D.  and  alii,  1999  : A comparative  analysis  of  transitions  from  education  to  work  in 
Europe  (CATEWE).  Volume  2.  Economic  and  Social  Research  Institute,  Working  Paper 
n°l  18(b),  Dublin. 

Jovanovic  B.,  1979  : “Job  Matching  and  the  Theory  of  Turnover”,  Journal  of  Political 
Economy  87. 

Muller  W.  and  Y.  Shavit,  1998  : From  School  to  Work.  A comparative  study  of  educational 
qualifications  and  occupational  destinations.  Oxford:  Clarendon  Press. 

Rose  J.,  1998  : Les  jeunes  face  a l ’emploi,  Desclee  de  Brouwer,  Paris. 

Ryan  P.,  1999  : "The  school-to-work  transition:  a cross-national  perspective",  working  paper, 
University  of  Cambridge,  September. 

Vincens  J.,  1997  : “ L’insertion  professionnelle  des  jeunes,  A la  recherche  d’une  definition 
conventionnelle  ”,  Formation/ Emploi,  N°60,  pp  21-36. 

Vincens  J.,  1998  : “ L’insertion  professionnelle  des  jeunes,  Quelques  reflexions  theoriques  ”, 
Formation/ Emploi,  N°61,pp  59-72. 


Appendix  A:  the  ‘Juniors’  as  a proxy  for  ‘new  entrants’  category 


To  construct  a proxy  for  a new  entrant  category  in  EULFS,  data  are  available  on  individuals' 
position  relative  to  the  ETS.  First,  the  highest  levels  of  general  education  and  post-academic 
training  attained  are  collected,  as  well  as  current  situation  concerning  education  and  training. 
By  drawing  on  the  national  educational  contexts,  we  can  try  to  reconstitute  the  typical  ages  of 
certification  at  the  end  of  the  main  training  routes  existing  in  national  educational  systems. 
The  quality  of  the  match  between  typical  and  real  age  depends  on: 

- The  precision  of  the  information  available  on  training  programmes.  The  more  the 
programmes  are  differentiated,  the  more  precisely  leaving  ages  can  be  attributed; 

- The  dissimilarity  of  behaviour  among  the  individuals  enrolled  in  these  programmes  in  terms 
of  passing  exams,  repeating  a year,  changing  orientations  and  so  forth; 

- The  multiplicity  of  possible  routes  amongst  the  different  tracks  and  levels  of  training,  which 
determines  the  number  of  possible  paths  to  reach  an  exit  point; 

- The  importance  of  going  back  to  training  and,  more  precisely,  the  ability  to  distinguish 
initial  education  and  training  from  post-initial  training. 

There  are  numerous  detailed  sources  of  information  on  the  organisation  of  training 
programmes  in  the  different  countries  of  the  European  Union  (cf.  Eurydice  database6).  To 
establish  typical  ages  of  certification,  the  OECD  has  analysed  in  each  country  about  ten 
different  programmes  (OECD  1998).  Using  this  information,  one  can  calculate  the  typical  age 
of  certification  for  individuals  in  the  European  Labour  Force  Surveys.  We  assume  that  the 
certification  age  corresponds  to  the  leaving  age  of  the  ETS7.  The  information  is  crossed  with 
the  observed  age  in  order  to  calculate  the  expected  length  of  time  elapsed  since  the  diploma 
was  obtained.  Different  categories  of  individuals  may  thus  be  identified.  First  of  all,  the 
'trainees',  which  include  individuals  who  are  still  in  a training  program  depending  on  the  ETS 
(including  altemance  training  programmes  with  the  training  course  taking  place  in  an 
establishment  of  the  ETS),  whether  or  not  they  are  in  the  labour  force  or  in  employment8. 

6 At  www.Euridyce.org 

An  assumption  that  doesn’t  take  into  account  the  issue  of  drop  out. 

8 with  a restriction  on  age  : from  15  to  35  years-old. 


Second,  the  'juniors',  who  represent  individuals  who  obtained  their  diploma  less  than  5 years 
previously.9  Third,  the  category  of  ' seniors' , sometimes  split  in  sub-categories,  which 
corresponds  to  individuals  whose  diploma  was  achieved  more  than  five  years  previously. 

This  leads  to  a distinction  between  two  categories  of  young  people  with  specific  statuses  in 
the  school-to-work  transition  process:  first,  the  trainees,  then  the  juniors.  We  will  focus  here 
on  the  latter  as  the  previous  is  treated  elsewhere. 


Table  1 : Typical  ages  of  leaving  the  ETS  according  to  levels  of  diploma,  based  on  typical  ages  of 
preparation  to  the  final  certificate,  collected  by  OECD 


ISCED 

0-1 

ISCED 

2 

I 

by  s 

general 

education 

SCED  3 (upi 

ingle  course  i 

school-based 

vocational 

training 

)er  secondary 

n ... 

alternated 

vocational 

training 

’) 

par  double 
course 
in  general 
education  and 
vocational 
training  (a) 

ISCED  5 

higher  non 
university 
diploma 

ISCED  6 

university 

diploma 

ISCED  7 

post 

university 

diploma 

‘OLM-type* 

countries 

Denmark 

16 

19 

20 

21 

21 

23 

24 

26 

31 

Germany 

18 

18 

19 

19 

19 

22 

21 

26 

28 

Netherlands 

18 

18 

19 

19 

20 

20 

- 

24 

27 

Austria 

15 

17 

18 

18 

19 

19 

21 

24 

26 

Southern 

countries 

Greece 

15 

18 

19 

19 

- 

- 

21 

23 

27 

Spain 

16 

17 

18 

17 

18 

19 

20 

22 

27 

Italy 

15 

18 

19 

18 

- 

19 

21 

23 

25 

Portugal 

15 

16 

17 

18 

18 

18 

22 

23 

26 

Other 

countries 

Belgium 

18 

18 

18 

19 

19 

- 

22 

23 

27 

France 

16 

17 

18 

19 

19 

20 

21 

21 

26 

Ireland 

15 

17 

18 

18 

18 

19 

20 

22 

24 

Luxembourg 

15 

18 

19 

19 

19 

- 

22* 

23* 

26* 

Finland 

16 

18 

19 

19 

19 

21 

23 

24 

28 

Sweden 

16 

18 

19 

19 

- 

- 

21 

23 

27 

U-Kingdom 

16 

17 

18 

18 

18 

- 

20 

21 

24 

Source  : inspired  from  Education  at  a glance  OECD,  Annexe  3. 


(a)  school-based  or  alternated 

* : estimation  based  on  neighbouring  countries  values  as  students  attend  higher  education  out  of  Luxembourg. 


9 Thus,  their  age  cannot  exceed  the  maximum  typical  age  by  more  than  5 years.  Again  there  is  a restriction  on  age,  from  15 
to  35  years-old. 


2.58 


231 


As  already  mentioned  in  the  second  chapter,  the  structure  of  educational  attainment  among 
juniors  varies  strongly  across  EU  countries.  Almost  every  kind  of  profiles  exists,  from  Spain, 
Portugal  and  Italy  which  have  dominant  ISCED  0-2  juniors  to  Belgium  where  the  junior 
group  with  ISCED  5-7  is  the  largest  group.  Between  them,  Germany,  Denmark,  Austria, 
Finland  and  Sweden  have  dominant  ISCED  3 levels.  As  a result,  the  average  age  of  juniors  at 
the  time  of  the  survey  differs  from  one  country  to  another.  It  is  low  for  Portugal  (17.1  years) 
and  goes  up  for  Denmark  (22.2  years). 


Table  2 - Juniors  by  level  of  education  and  training  attained  and  their 
average  age  at  the  time  of  leaving  ETS 


Juniors 

Juniors  average 

ISCED  0-2  (%) 

ISCED  3 (%) 

ISCED  5-7  (%) 

Total  (%) 

age  at  the  time  of 
leaving  ETS  (years) 

‘OLM-type’  countries 

Denmark 

19 

53 

28 

100 

22.2 

Germany 

15 

57 

28 

100 

20.8 

Netherlands 

23 

47 

29 

100 

19.6 

Austria 

16 

75 

9 

100 

18.8 

Southern  countries 

Greece 

26 

53 

21 

100 

19.1 

Spain 

47 

16 

37 

100 

19.3 

Italy 

52 

41 

7 

100 

18.3 

Portugal 
Other  countries 

63 

17 

20 

100 

17.1 

Belgium 

23 

36 

42 

100 

19.7 

France 

18 

46 

36 

100 

20.2 

Ireland 

25 

39 

36 

100 

18.7 

Luxembourg 

36 

36 

28 

100 

19.2 

Finland 

17 

52 

31 

100 

20.9 

Sweden 

18 

62 

19 

100 

19.2 

U-Kingdom 

39 

37 

25 

100 

18.2 

EU 

32 

43 

25 

100 

19.4 

Source:  Eurostat  - CLFS,  1995  to  1997 


O 


2.59 


January  2001 


Prepared  as  part  of  the  TSER  project: 

Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe 


New  Entrants  and  Experienced  Workers  on 
European  Labour  Markets 


Thomas  Couppie 
Cereq 

10  place  de  la  Joliette 
F-13474  Marseille  Cedex  02 
e-mail:  couppie@cereq.fr 

Michele  Mansuy 
Insee 

17  rue  Menpenti 

F-13387  Marseille  Cedex  10 

e-mail:  michele.mansuy@insee.fr 


233 


WORKING  PAPERS 


Introduction 


Compared  to  the  rest  of  the  labour  force  (termed  'seniors')  new  entrants  (also  termed  'juniors') 
are  defined  by  their  lack  of  labour  market  experience.  The  transition  period  thus  corresponds 
to  acquiring  experience  through  a first  job.  For  new  entrants,  this  first  experience  consists  in 
developing  skills  but  also  in  adapting  themselves  to  company  rules.  However,  the  forms  of 
this  transition  process  vary  considerably  from  one  country  to  another  (see  Ryan,  1999). 

Our  purpose  here  is  to  see  how  the  labour  demand  determination  affects  young  people's 
integration  into  the  labour  market.  Do  companies  differentiate  between  new  entrants  and  other 
manpower  categories?  To  what  extent  are  new  entrants  disadvantaged? 

There  have  already  been  a number  of  studies  comparing  national  systems  of  training  and 
employment  and  their  consequences  in  terms  of  access  to  different  social  categories  (Muller 
and  Shavit,  1998,  is  a reference  book).  In  the  present  paper,  emphasis  is  placed  on  the 
determinants  of  labour  demand.  This  involves  understanding  the  outcomes  produced  by  the 
hiring  policies  of  employers.  In  particular,  the  central  question  is  the  room  that  companies  are 
ready  to  make  for  new  entrants  amongst  the  other  labour- force  categories.  The  answer  is 
complex.  The  idea  of  a single  model  of  behaviour,  where  companies  act  as  simple  economic 
agents  according  to  the  canonical  rules  of  perfect  competition  has  long  been  called  into 
question.  Theories  of  segmentation  argue  that  in  the  context  of  markets  the  model  of  perfect 
competition  is  not  the  point  of  equilibrium  (see  Doeringer  and  Piore  1971,  as  the  reference 
work).  Following  the  initial  work  of  Maurice,  Sellier  and  Silvestre  (1982),  Marsden  (1986) 
provided  a more  institutional  and  macro-economic  version  of  labour  market  segmentation 
including  the  ‘societal  effect’  analysis.  In  such  a framework,  both  the  production  of 
qualifications  and  the  reproduction  of  the  labour  force  depend  strongly  on  the  overall 
organisation  of  industrial  relations  and  on  the  structure  of  labour  markets.  The  transition 
process  would  therefore  reflect  the  way  of  regulating  young  people’s  entry  into  the  labour 
market  rather  than  being  the  sole  consequence  of  individual  characteristics. 

According  to  Garonna  and  Ryan  (1989),  there  are  different  systems  regulating  the  entry  of 
youth  in  Europe.  Labour-market  organisation,  symbolised  by  the  trio  internal 
market/occupational  market/unorganised  or  casual  market,  is  a key  factor  in  youth  integration. 
But  besides,  the  state  of  industrial  relations  and  the  linkages  between  ETS  and  employers  are 
also  crucial. 


ERjt 


3.1 


97A 


Garonna  and  Ryan’s  presentation  possesses  three  main  advantages: 

First,  collective  agreements,  custom  and  practice  at  the  workplace  are  part  of  the  picture. 
Thus  insiders'  and  outsiders'  relative  positions  become  clearer.  Among  other  authors, 
Lindbeck  and  Snower  (1988)  studied  how  insiders  obtain  higher  wages  and  more  secure  jobs 
by  collective  bargaining.  But  they  assume  that  the  insiders/outsiders  dichotomy  is  an 
important  cause  of  unemployment,  and  further,  of  unemployment  persistence1.  Empirical 
observation  does  not  support  this  assumption.  On  the  other  hand,  Cahuc  and  Zylberberg 
( 1 996)  have  a different  interpretation:  they  consider  that  insiders  may  negotiate  their  wages, 
but  also  specific  wages  for  newcomers  or  layoff  premiums.  In  that  case,  insiders'/outsiders' 
relative  positions  produce  discrimination  (against  outsiders)  rather  than  unemployment.  This 
latter  version  of  the  insiders/outsiders  dichotomy  will  be  used  here. 

Second,  ETS  and  productive  systems  linkages  are  included,  and  their  outcomes  are  developed 
(For  instance,  in  OLM  countries,  firms  are  strongly  involved  in  initial  vocational  training.  But 
on  the  other  hand  LM  structure  is  not  sufficient  to  characterise  the  transition  process:  this 
process  is  smoother  in  Japan  than  in  any  other  country  with  dominant  internal  labour  markets, 
because  firms  there  hire  many  school  leavers,  relying  on  schools  to  select  the  applicants). 

Third,  it  takes  into  account  market  regulation  changes,  such  as  deregulation  and  search  for 
flexibility.  For  more  than  twenty  years,  international  competition  and  persistently  high 
unemployment  have  put  into  question  the  former  regulatory  systems2.  Youth  were  a central 
concern  of  deregulatory  policies,  aimed  at  favouring  their  inclusion  into  the  workplace.  These 
changes  are  included  in  Garonna  and  Ryan’s  models. 

Garonna  and  Ryan  distinguish  three  ideal-typical  youth  regulatory  systems:  regulated 
inclusion,  selective  exclusion  and  competitive  regulation. 


1 -Insiders  obtain  higher  wages  than  outsiders’  reservation  wage.  Newcomers  are  hired  at  the  same  wage  level  as  insiders. 
The  smaller  the  insiders  group,  the  higher  the  wage  level.  According  to  this,  insiders'  bargaining  produces 
unemployment.  As  insiders  only  participate  in  collective  bargaining,  wages  remain  high  even  if  the  unemployment  level 
increases,  and  outsiders  remain  unemployed.  So  insiders’  position  also  produces  unemployment  persistence. 

2 Regulatory  systems  are  here  closer  to  the  French  regulation  school  than  to  the  use  of  the  word  regulation  in  the  US  or  UK 
economic  literature.  However,  Garonna  and  Ryan  do  not  share  all  the  regulation  school  framework  : they  put  emphasis 
on  differences  within  economies  as  well  as  between  them,  and  on  industrial  relations  more  than  the  organisation  of  work 
and  education  (see  Garonna  and  Ryan,  1991,  p.77,  note  4).  In  a recent  paper,  Boyer  and  Juillard  (1998)  forecast  three 
major  wage  labour  nexus  in  Europe  for  the  XXI  century.  The  similarity  between  the  three  forms  they  define  and  the 
Garonna  and  Ryan  youth  regulatory  systems  is  striking. 


0 

ERLC 


!*>  ( ■ 


3.2 


235 


Regulated  inclusion  in  the  context  of  dominant  occupational  markets 

Here,  the  social  partners  monitor  the  training  system  directly.  The  most  notorious  example  is 
the  German  case  (Mehaut  1993).  The  gains  linked  to  a higher  level  of  skills  of  their 
workforces,  without  the  threat  of  poaching,  is  the  main  motive  of  employers’  goal  and 
strategy.  On  the  other  hand,  unions  seek  ways  to  reproduce  occupational  communities  while 
regulating  youth  activity  in  the  workplace.  Collective  bargaining  includes  training 
codification  and  a minimum  standard  of  training  quality  under  public  administration  control. 
Employers  design  skilled  jobs  in  order  to  use  standardised  skills  provided  by  the  ET  system 
efficiently.  The  occupational  qualification  needed  to  obtain  skilled  positions  on  occupational 
markets  is  acquired  through  a structured  system  of  altemance,  the  most  traditional  form  of 
which  is  apprenticeship.  Following  this  period  of  altemance,  young  workers  are  entitled  to 
enter  the  related  occupational  market.  Any  target  occupation  or  any  kind  of  firm  is  open  to 
them  within  this  occupational  market.  For  a given  kind  of  vocational  training  obtained,  it 
should  thus  be  possible  to  observe  the  same  diversity  of  jobs  for  new  entrants  as  for  more 
experienced  workers  and  a strong  homogeneity  in  their  job  profiles.  Thus,  work  experience 
has  a limited  impact  on  recruitment  decisions.  In  this  kind  of  regulation,  skills  are  transferable 
between  companies  but  adults  are  partly  protected  from  young  people’s  competition,  because 
access  to  occupations  is  regulated  by  the  completion  of  the  corresponding  qualification.  So 
skilled  adults  only  compete  with  skilled  new  entrants  applying  for  the  same  occupation. 

Selective  exclusion  in  the  context  of  dominant  internal  markets 

Internal  markets  in  the  primary  segment  are  characterised  by  skilled  jobs,  high  wage  levels 
and  restricted  access.  They  correspond  to  specific  job  or  company  profiles.  In  the  typical  form 
of  the  internal  market,  the  qualification  is  obtained  within  the  company  and  is  not 
transferable,.  To  achieve  high  profitability  in  the  long  run  by  retaining  skilled  workers, 
companies  favour  internal  promotion  and  give  a significant  bonus  to  seniority.  Here,  the 
balance  between  the  Taylorist  tendency  to  reduce  skill  requirements  and  the  interest  in  the 
technical  benefits  of  high  employee’s  competence  dominates  employers’  goals  and  strategy, 
where  the  main  goal  for  unions  is  to  secure  pay  and  jobs  for  settled  workers.  Wages  are  fixed 
according  to  the  present  job  position  and  not  to  the  individual  qualifications  as  they  are  in  the 
regulated  inclusion  model.  This  form  of  regulation  where  appropriate  skills  are  obtained 
within  the  company  leads  to  increasing  wages  with  seniority. 


Here,  insiders  protect  themselves  against  outsiders.  When  unions  are  strong,  even  if  outsiders’ 
reservation  wage  is  lower  than  that  of  insiders,  the  latter  may  bargain  on  hiring  and  firing 


O 

ERIC 


236 


3.3 


costs,  as  well  on  outsiders’  wages,  in  order  to  preserve  their  rewards3.  Conversely,  new 
entrants  are  likely  to  be  recruited  at  the  lowest  skill  levels.  Presumed  less  productive  than 
their  seniors  because  of  their  inexperience,  new  entrants  are  thus  placed  at  the  end  of  the 
queue  for  hiring. 

On  the  other  hand,  the  companies  are  themselves  more  concentrated  in  one  segment  of  the 
labour  market  on  the  basis  of  their  size  and  their  activity.  Thus,  large  companies  are  often 
more  attractive  because  they  can  more  easily  develop  an  internal  segment  fed  by  in-house 
promotion  than  small-scale  companies  offering  few  prospects  of  promotion.  Furthermore, 
economic  activities  are  sorted  from  the  most  attractive  to  the  least  attractive  according  to  the 
level  of  compensation  they  offer.  However,  the  development  of  national  productive  apparatus 
-the  relocation  of  certain  manufacturing  activities,  the  growth  of  the  services  sector,  the 
increasing  use  of  the  new  information  and  communication  technologies-  is  leading  to  a 
shrinking  of  internal  markets.  In  that  context,  it  is  even  more  difficult  for  new  entrants  to  enter 
the  internal  segments. 

To  sum  up,  in  such  an  organisation  of  the  labour  market,  companies  accepting  new  entrants 
differ  from  companies  employing  their  seniors.  Internal  markets  are  often  closed  to  juniors. 
More  often  integrated  into  the  segments  of  the  casual  labour  market,  young  people  are  more 
likely  to  hold  unstable  jobs  with  a high  turnover  rate.  Nevertheless,  internal  labour  markets  do 
not  systematically  produce  selective  exclusion:  a strong  position  of  insiders,  protecting 
themselves  from  outsiders’  competition,  is  also  a necessary  condition.  It  is  not  the  case  in 
Japan,  as  stressed  by  Ryan,  1999.4 

Competitive  regulation  in  the  context  of  dominant  competitive  markets 

In  this  kind  of  regulation,  under  pressure  from  the  economic  environment  (strong  competition, 
an  uncertain  production  climate)  or  due  to  an  imbalance  between  supply  and  demand  on 
labour  markets,  firms  aim  at  making  the  labour  factor  profitable  in  the  short  run.  This  can 
only  occur  in  the  context  of  high  unemployment  rates  and  of  weak  union  bargaining  power. 
Companies  seek  maximum  productivity  from  the  labour  factor  and  attempt  to  minimise 
associated  costs  (whether  direct  or  indirect).  Insiders  have  a limited  bargain  power  and  cannot 
protect  themselves  from  outsiders’  competition.  Employers  are  strong  enough  to  put  pressure 


e.g.  by  imposing  minimum  wages. 

In  Japan,  where  ILMs  prevail,  the  linkage  between  schools  and  companies  is  strong,  and  Japanese  firms  readily  recruit 
school-leavers. 


on  wages  and  to  weaken  labour  market  control  (e.g.  by  suppressing  restrictive  rules  for  hires 
and  layoffs).  From  this  perspective,  the  firm  acts  as  a consumer  of  skills  and  gives  up  any  role 
in  their  production.  The  firms  then  seek  skills  on  the  labour  market  at  the  lowest  cost  by 
effecting  trade-offs  between  accumulated  work  experience  and  level  of  training  attained.  In 
the  presence  of  a regulation  of  the  minimum  wage  level,  this  leads  to  a process  of  excluding 
the  least  qualified  new  entrants,  at  the  same  time  that  competition  emerges  between  graduated 
new  entrants  and  lower  qualified  but  experienced  workers.  This  type  of  regulation,  with  high 
responsiveness  of  pay  levels  to  economic  conditions,  is  not  necessary  restricted  to  the  lowest 
levels  of  skills,  but  may  apply  also  to  high  levels  of  transferable  skills.  In  the  most  skilled 
jobs,  such  as  professionals,  younger  people  at  a lower  labour  cost  may  even  replace  the  most 
experienced  and  highly  paid  employees. 

Some  national  economies  show  a prevalence  of  one  of  the  three  forms  above.  It  is  the  case  in 
Germany,  with  the  regulated  inclusion  model.  Other  countries  are  more  mixed,  combining 
two  or  three  types.  In  particular,  the  third  model  is  only  a tendency  and  has  never  applied  on  a 
national  scale  in  recent  years5.  Therefore,  it  is  a convenient  way  to  depict  the  recent  evolution 
of  former  regulatory  systems.  In  the  following,  we  will  consider  the  three  most  realistic 
national  combinations: 

1 - Dominant  regulated  inclusion  (RI),  as  described  above. 

2 - Dominant  selective  exclusion  (SE) , as  described  above. 

3 - selective  exclusion  mixed  with  competitive  regulation  (SE+CR) 

This  third  model  is  an  evolution  from  the  selective  exclusion  model.  It  takes  place  in  the 
framework  of  dominant  internal  markets,  when  deregulation  and  flexibility  policies  are 
introduced  under  economic  pressure.  Apart  from  wage  flexibility,  two  factors  of  evolution 
from  the  former  model  may  be  noticed. 

First,  the  sharply  increasing  level  of  education  of  ETS  leavers  leads  to  a disruption  of  earlier 
rules  for  acquiring  vocational  training  within  internal  markets,  which  has  now  basically 
reverted  to  the  initiative  of  educational  institutions,  public  or  private. 

Secondly,  public  intervention  in  the  labour  market  may  lead  to  introduce  a greater  flexibility 
of  wages  and  labour  contracts.  Firms  may  use  public  integration  schemes  to  lower  their  direct 


233 

3.5 


or  indirect  labour  costs.  As  a consequence,  new  entrants  will  be  less  disadvantaged  in  their 
access  to  jobs. 

Using  the  three  previous  combinations,  we  attempt  to  bring  out  some  specific  features  of  new 
entrants'  work  situation  on  the  labour  market  in  terms  of  access  to  jobs,  type  of  jobs  they  hold 
(identified  by  the  kind  of  hiring  company,  the  occupation,  the  labour  contract). 


Empirical  evidence  of  regulated  entry  patterns  for  new  entrants 

Pointing  out  distinctive  characteristics  of  new  entrants  on  labour  markets  will  allow  us  to 
identify  empirically  regulated  entry  patterns.  First,  we  translate  the  three  combinations  of 
Garonna  and  Ryan  ideal  types  into  assumptions  about  contrasting  positions  on  labour  markets 
for  new  entrants  and  experienced  workers.  We  then  briefly  present  the  data  available  for  our 
study  and  the  empirical  indicators  corresponding  to  our  assumptions.  Finally,  we  discuss  the 
clusters  produced  by  empirical  results. 

Assumptions  on  entry  patterns  based  on  regulatory  systems  of  youth  entry 
and  their  expected  influence  on  the  school  to  work  transition 


Here  are  analysed  the  different  regulatory  systems  previously  presented  and  how  they  should 
influence  the  school  to  work  transition  and  contribute  to  shape  pathways  on  labour  markets. 
The  purpose  now  is  to  identify  the  predictable  specific  features  of  school  to  work  transition 
induced  by  each  of  the  regulatory  systems  and  to  highlight  the  way  these  features  combine 
together. 


Initial  Education  and  Training  System 

The  importance  and  the  role  of  initial  education  and  training  vary  strongly  from  one 
regulatory  system  to  another.  This  can  be  linked  to  the  constituent  pattern  of  skills  that  are 
recognised  and  used  by  firms.  The  level  of  recognition  of  diplomas  and  titles  delivered  by  the 


5 


Garonna  and  Ryan  mention  an  article  of  Gollan  (1937)  describing  such  a model  as  prevailing  in  the  UK  after  the  economic 
crisis  of  1929. 


3.6 


PQO 


ETS  varies  on  the  labour  market  and  can  be  ranked  from  a standardised  access  rule  function 
to  a more  elusive  signal  value  (Verdier  and  Mobus  1999). 

Vocational  qualification  standards,  elaborated  and  shared  by  all  the  economic  agents  (firms, 
unions,  State)  are  a component  of  the  regulated  inclusion  system.  These  standards  constitute 
access  rules  to  the  different  occupational  markets.  In  such  a system,  the  definition  of 
individual  qualification  is  close  to  the  content  of  skills  required  by  companies.  So,  vocational 
education  and  training  is  widespread.  At  the  same  time,  youth  qualifications  also  include,  as 
well  as  formal  and  practical  knowledge,  preparation  for  the  working  world  and  life  within 
firms.  That  preparation  can  be  formally  linked  with  vocational  training  (as  in  apprenticeship) 
or  not.  Anyway,  it  leads  to  a high  percentage  of  young  people  combining  education  and 
employment. 

In  contrast,  a selective  exclusion  system  uses  qualifications  more  as  an  information  signal  on 
individual  abilities  than  as  an  actual  description  of  usable  skills.  This  signal  stands  for  a level 
of  education  rather  than  a specific  training  content.  As  a consequence,  vocational  education 
and  training  is  less  valued  by  firms  and  students  than  in  a RI  model.  Access  to  skilled 
positions  being  regulated  by  firm-specific  experience  and  training,  young  people  with  early 
working  experience  are  not  in  a stronger  position.  Weak  valuation  of  a first  job  experience, 
limited  job  opportunities  in  low  skilled  positions  and  prevalence  of  education  attainment,  all 
sharply  reduce  incentives  to  get  a job  during  the  course  of  studies. 

In  the  case  of  SE+CR,  firms  tend  to  reduce  labour  costs  and  favour  short-term  profitability 
(Ashton  1994).  Therefore,  they  tend  to  limit  their  participation  to  the  production  of  skills, 
upgrading  their  interest  in  both  general  and  vocational  education.  In  such  a system,  young 
trainees  of  the  IETS  with  working  experience  will  have  an  advantage  in  competing  with  both 
more  experienced  workers  and  inexperienced  trainees.  This  would  lead  to  the  development  of 
double  statuses  among  young  trainees. 

Thus,  one  can  expect  to  observe:  numerous  connections  between  initial  training  and 
employment  in  regulated  inclusion,  few  connections  in  selective  exclusion  systems,  and 
intermediate  to  high  level  of  double  statuses  in  SE+CR  cases. 


3.7 


Access  to  jobs  and  job  mobility 

Whatever  the  regulatory  system,  new  entrants  are  outsiders  on  the  labour  markets,  at  the  end 
of  their  participation  in  the  ETS.  This  is  obviously  the  case  in  countries  where  participation  in 
training  courses  is  clearly  separated  from  participation  in  labour  markets  (as  in  Italy).  But  this 
is  also  true  in  the  countries  where  the  two  situations  overlap,  either  in  apprenticeship  (as  in 
German  dual  system)  or  mainly  outside  this  frame  (as  in  the  Netherlands).  Even  in  the  latter 
countries,  leavers  from  school  or  apprenticeship  experience  a transition  to  a standard  worker 
status.  In  a context  of  jobs  rationing,  one  would  expect  that  the  transition  process  would  result 
in  unemployment  - at  least  transitory  - among  new  entrants.  Therefore,  new  entrants  are 
more  exposed  to  unemployment  risk  than  experienced  workers.  But  the  gap  between  the  two 
groups  is  linked  with  the  corresponding  regulatory  system  of  youth  entry.  In  a regulated 
inclusion  frame,  the  unemployment  gap  should  be  quickly  reduced,  as  skilled  new  entrants 
fulfil  the  entry  criteria  for  occupational  labour  markets.  As  a result,  they  are  weakly 
differentiated  from  more  experienced  workers  by  employers.  The  matching  process  is  thus 
reduced,  and  juniors  are  not  affected  by  prolonged  unemployment  spells.  Skills  being 
transferable,  job  mobility  is  rather  high,  among  experienced  workers  as  well  as  among  new 
entrants. 

On  the  other  hand,  in  a selective  exclusion  entry  process,  employers  hire  experienced  workers 
first  among  job  applicants.  In  a context  of  high  unemployment,  this  can  create  persistent 
difficulty  in  job  access  for  juniors,  and  result  in  a specific  pool  of  unemployed  juniors 
engaged  in  long-term  job  search.  Moreover,  job  access  difficulties  tend  to  reduce  incentives 
for  job  mobility  among  settled  workers,  either  juniors  or  seniors,  whatever  the  quality  of  the 
job  held. 

In  a SE+CR  model,  employers  intend  to  develop  flexibility  in  manpower  management 
(Garonna  and  Ryan  1991).  From  that  perspective,  employers  promote  a regular  group  of 
precarious  workers  in  order  to  manage  numerical  flexibility  (Marsden  1992).  That  group  also 
acts  as  a lobbying  instrument  that  affects  job  security  among  settled  manpower  and  reduces 
their  bargaining  power  (Marsden  1991,  Ryan  1991).  New  entrants  take  an  important  place  in 
such  a pool  of  workers.  Competition  among  new  entrants  is  open,  and  firms  choose  among 
them  according  to  their  level  of  education  and  their  expected  total  labour  cost.  Having  lower 
pay  demands  and  being  less  reluctant  to  enter  precarious  work  relationships,  new  entrants  can 
easily  compete  with  more  experienced  outsiders  too.  As  a consequence,  job  turnover  is  high, 
and  recurrent  short-lasting  unemployment  spells  are  common,  both  for  new  entrants  and 
experienced  outsiders.  However,  insiders  remaining  numerous  among  experienced  workers, 


we  can  find  in  this  model  a higher  level  of  job  mobility  among  new  entrants.  On  the  other 
hand,  the  gap  in  unemployment  rates  between  new  entrants  and  experienced  workers  is 
expected  to  be  much  narrower  than  in  the  SE  case. 


Job  dispersion  of  new  entrants  by  industry 

Substantial  and  persistent  differences  in  earnings  and  skill  structure  are  observed  between 
industries.  Are  the  relative  variations  in  the  youth  share  by  industry  related  to  the  different 
patterns  of  youth  entry  regulation  on  the  labour  market?  Indeed,  as  a general  feature,  one  can 
expect  that  high  wage  firms  and  activities  select  already  experienced  workers  as  a more 
productive  group  than  new  entrants,  operating  as  in  Thurow’s  (1975)  job  competition  model 
on  the  labour  market.  However,  this  general  feature  should  be  affected  by  the  dominant  youth 
regulatory  system. 

In  a regulated  inclusion  system,  the  variation  of  new  entrants’  share  among  economic 
activities  is  limited.  Firstly,  the  qualification  process  within  the  workplace  already  provides 
work  experience,  potentially  linked  with  targeted  skilled  positions.  Secondly,  the 
qualification-based  criterion  overturns  the  experience-based  criterion  to  favour  skilled  job 
entry  for  juniors. 

At  the  opposite,  selective  exclusion  favours  youth  concentration  in  low  skilled  jobs  and  firms 
that  belong  to  the  secondary  segment  (Marsden  and  Ryan,  1990).  Being  outsiders,  new 
entrants  are  disadvantaged.  Many  of  them  can  only  access  low  skill  job  positions  or  less 
attractive  firms.  Internal  market  job  positions  are  mainly  occupied  by  insiders. 

In  a SE+CR  system,  insiders'  bargaining  power  is  reduced  and  firms  seek  to  ‘open’  their 
recruitment  to  varied  groups  of  workers.  As  juniors  have  a higher  level  of  education,  this 
might  counterbalance  their  lack  of  experience.  This  would  more  likely  happen  if  reduced  total 
labour  costs  are  involved.  It  is  the  case  when  on-the-job  training  costs  are  low.  It  is  also  true 
when  hiring  a junior  is  associated  with  a downgrading  in  the  returns  to  completed  training. 
New  entrants’  employment  space  is  consequently  more  open  than  in  a SE  model. 


242 


3.9 


Returns  to  qualifications  and  experience 

The  outcomes  that  new  entrants  can  expect  to  achieve  in  the  labour  market  vary  according  to 
their  ability  to  attain  a good  job.  Education  and  training  received  is  a major  component  of 
these  abilities.  As  proposed  by  Becker  (1993),  investment  in  human  capital  is  expected  to 
produce  subsequent  returns  on  the  labour  market  (wage  attained,  career  prospects,  prestige  of 
jobs). 

Other  theories  see  the  economic  role  of  the  diploma  in  a very  different  way.  For  Arrow 
(1973),  the  diploma  acts  like  a filter,  revealing  the  applicant’s  performance  ability,  rather  than 
proving  accumulation  of  general  knowledge  and  skills.  For  young  people,  education  and 
training  are  an  information  investment,  rather  than  a human  capital  accumulation.  In  an 
extreme  version,  education  and  training  curricula  don’t  matter.  Spence  (1974)  offers  another 
version  of  screening  processes  related  to  hiring.  In  his  market  signalling  theory,  employers 
use  diplomas  as  a signal  of  potential  productivity.  Here,  the  diploma  may  or  may  not  improve 
potential  productivity.  The  cost  of  education  and  training  is  supposed  to  be  dissuasive  for  less 
competent  persons.  On  the  contrary,  the  supplementary  gain  attached  to  qualifications  is 
supposed  to  be  profitable  for  competent  persons.  In  that  case,  employer  and  job  applicant 
expectations  turn  out  to  be  right. 

All  these  models  assume  that  wages  are  fixed  according  to  the  (expected)  productivity.  We 
assume  here  that  pay  structure  and  career  prospects  don’t  depend  only  on  individual 
rationality  but  more  on  the  state  of  industrial  relations  and  the  labour  organisation. 

Indeed,  the  three  regulatory  models  of  youth  entry  are  related  to  very  different  patterns  of 
education  and  training  rewards. 

In  the  ideal-type  of  occupational  arrangements  linked  with  regulated  inclusion , wages  mainly 
reward  the  qualifications  held  by  workers  and  not  their  work  experience  (Marsden  and  Germe 
1991).  The  wage  progression  curve  in  the  course  of  the  career  is  rather  flat,  as  long  as  the 
qualifications  held  are  not  upgraded.  As  a consequence,  the  skill  level  of  jobs  offered  to  new 
entrants  is  adjusted  to  their  qualifications  (there  is  no  downgrading)  and  juniors'  pay  levels  do 
not  differ  very  much  from  those  of  seniors. 


On  the  contrary,  selective  exclusion  leads  to  differing  consequences  for  the  structure  of 
education  and  training  rewards.  First,  in  the  primary  segment,  internal  arrangements  favour  an 
increasing  return  to  seniority  within  firms,  as  a way  to  reward  on-the-job  training  and  specific 
skills  acquired  through  experience.  It  includes  possible  internal  promotions  to  skilled 


3.10 


243 


positions.  The  counterpart  is  a reduced  return  to  training  when  entering  a new  firm  and  a 
subsequent  transitory  downgrading.  If  job  competition  is  intense  and  insiders  in  a strong 
position,  even  qualified  new  entrants  may  enter  low  skilled  jobs  in  the  secondary  segment. 
Hence,  occupational  downgrading  of  qualified  young  workers  into  lower  skilled  positions  is 
fairly  extensive.  Later  upgrading  is  possible,  as  a reward  for  seniority.  Thus,  return  to  training 
should  be  contrasted  for  new  entrants  and  experienced  workers. 

In  the  case  of  SE+Ci?  model,  firms  tend  to  limit  the  reward  for  experience  and  the  reward  for 
qualifications.  It  is  achievable,  as  insiders'  bargaining  power  is  weaker  than  in  a pure  SE 
model.  This  leads  to  a downgrading  of  higher  qualified  new  entrants  into  lower  skilled 
positions,  at  least  during  the  transition  period.  As  a result,  for  new  entrants,  returns  to 
education  and  training  are  lower  than  for  experienced  workers.  However,  insiders’  position 
being  weaker,  the  gap  between  the  two  categories  is  expected  to  be  narrower  than  in  a SE 
model. 

Table  1 below  summarises  the  assumptions  we  made  and  recapitulates  the  expected  effects  on 
labour  market  outcomes  for  trainees  of  the  ETS  and  juniors. 


0 


ERIC 


3.11 


244 


Table  1:  Transition  characteristics  by  youth  regulatory  system 


Transition 

characteristics 

Regulated 

Inclusion(RI) 

Selective 

Exclusion(SE) 

SE  + CR6 

Education 

Learning  and 

High  percentage  of 

Low  percentage  of 

Medium  to  high  percentage  of 

working  double 
statuses 

learning  and 
working  double 
statuses. 

learning  and  working 
double  statuses 

learning  and  working  double 
statuses. 

Vocational 

education 

Extensive  and 
codified  vocational 
education 

Very  limited  codified 
vocational  education 

Codified  vocational  education 

Job  access 

Juniors  and  seniors 
similar  risks  of 
unemployment 

Juniors  have  a higher 
risk  of  unemployment 

Juniors  and  seniors 
unemployment  risks  are  closer 
than  in  SE  system. 

Job  mobility 

Job  mobility  for 

Medium 

limited 

Medium  to  high 

seniors 

Job  mobility  gap 
for  juniors 

Medium 

limited  by  incentives 
to  keep  jobs 

High 

Job 

No  discrimination 

Strong  discrimination 

Weaker  discrimination  than  in 

discrimination 

towards  juniors  by 
industry. 

on  primary  segment 

a pure  ‘SE’ 

Downgrading 

risk 

Limited  specific  risk 
For  juniors 

High  risk  for  juniors 

Medium  to  High  risk  for 
juniors 

’ CR  means  Competitive  Regulation 


o 

ERLC 


3.12 


Available  data  and  empirical  indicators 


The  empirical  application  in  the  present  paper  is  based  on  data  extracted  from  the  annual 
European  Labour  Force  Survey  between  1993  and  19977.  The  annual  survey  is  the 
compilation  of  a subset  of  variables  from  national  Labour  Force  Surveys  in  every  member 
States  (EUROSTAT  1996).  The  variables  are  regularly  revised  and  the  last  revision  occurred 
in  1998.  The  surveys  are  generally  conducted  on  large  samples,  the  sampling  rate  varying 
across  countries.  Although  the  sample  is  never  fully  renewed  from  one  year  to  another,  panel 
data  are  not  available  at  the  European  level.  Information  collected  is  almost  entirely  cross- 
sectional,  and  the  purpose  of  the  surveys  is  to  give  reliable  and  comparable  information  on  job 
structure  and  unemployment  across  European  Union.  To  do  this,  the  survey  promotes 
international  definitions  of  activity,  namely  definitions  elaborated  by  the  International  Labour 
Office. 

To  implement  the  two  groups  of  young  people  facing  the  school  to  work  transition,  we  used: 

- Current  participation  in  an  initial  educational  and  training  course, 

- Current  position  on  the  labour  market, 

- A proxy  of  the  age  of  leaving  IETS8  constructed  on  the  basis  of  three  sets  of  data:  highest 
levels  of  education  and  training  attained,  the  current  age  of  young  (both  data  present  in  the 
survey)  and  the  typical  graduation  ages  according  to  the  educational  level  attained  (OCDE 
1998). 

Table  2 summarises  the  different  indicators,  related  to  the  assumptions  we  previously  made. 
Seven  indicators  cover  the  extension  of  working  experiences  among  trainees,  the  place  of 
vocational  training,  unemployment  features,  job  mobility,  the  possible  concentration  into 
specific  sectors  and  the  occupational  downgrading  associated  with  youth  employment. 

To  sum  up,  each  regulatory  system  will  be  characterised  as  follows: 

In  a RI  system,  the  two  indicators  related  to  education  and  training  will  have  a high  value.  No 
difference,  or  a very  slightly  one,  is  expected  between  juniors'  and  seniors'  unemployment 
rates.  As  skills  are  transferable,  but  unions'  bargaining  limits  the  number  of  precarious  jobs, 
the  general  job  mobility  rate  is  likely  to  be  intermediate.  The  key  to  skilled  jobs  being 


7 The  data  has  kindly  been  provided  by  EUROSTAT  which  is  the  Statistical  Office  of  the  European  Union.  Of  course, 
EUROSTAT  is  neither  responsible  for  the  use  of  the  data,  the  interpretations  drawn  nor  the  views  expressed  by  the 
authors. 

8 not  available  in  the  data. 


ERIC 


3.13 


248 


qualification  and  not  experience  nor  seniority,  juniors  are  not  expected  to  be  much  more 
mobile  then  seniors. 


Table  2:  Empirical  indicators  used  • 

summary  of  their  content 

Learning  and  working  double  statuses 

Share  of  working  vocational  or  post-secondary 
trainees  (people  having  a job  among  those  attending 
post-secondary  education  or  vocational  training) 

Vocational  education 

Share  of  vocational  qualified  among  employed 
juniors9 

Job  access 

Ratio  of  unemployment  rates  (UNJ/UNS) 

Job  mobility  of  juniors 

Transition  rate  for  juniors  from  a job  held  the 
previous  year  to  another  employment  status 
(unemployed,  new  job) 

Job  mobility  of  seniors 

the  same  for  seniors  (same  definition  as  above) 

Job  discrimination 

Importance  of  variation  of  juniors’  share  among 
total  labour  force  by  industry  (the  standard  deviation 
is  divided  by  weighted  mean,  economic  activity  is 
divided  in  12  industries) 

Relative  downgrading  risk  : ISEI  scores  related  to 
the  highest  level  of  education  and  training  attained 

ISEI  scores  ratios  between  juniors  and  seniors  by 
level  of  education  and  training  attained  : (as  the 
weighted  mean  of  ratios  calculated  on  average  ISEI 
scores  in  two  different  groups  (one  of  juniors  and 
seniors  with  intermediate  level  of  education  and 
training  (ISCED  3)  and  one  of  juniors  and  seniors  who  . 
have  achieved  higher  education  (ISCED  5 to  7)) 

For  similar  reasons,  juniors  are  expected  to  be  present  in  all  industries,  without  discrimination 
(indicator  6).  Juniors  are  supposed  to  have  occupations  matching  their  qualification  level,  so 
their  ISEI  score  ratio,  relatively  to  seniors  is  expected  to  be  near  1 . 

In  a selective  exclusion  model,  the  two  indicators  related  to  education  and  training  are 
expected  to  have  low  values.  Unemployment  risk  is  likely  to  be  higher  for  juniors.  General 

9 Data  on  unemployed  or  out  of  the  labour  force  were  not  available. 


job  mobility  is  expected  to  be  low  (indicators  4 and  5).  Discrimination  towards  juniors  is 
high.  The  ISEI  score  ratio  for  juniors  is  likely  to  be  lower  than  for  seniors  with  the  same 
qualifications  (low  value  of  indicator  7). 

In  a SE+CR  model,  education  characteristics  are  different  from  a SE  model:  codified 
vocational  education  is  more  extended  and  double  statuses  are  fairly  frequent.  Juniors  still 
have  a higher  unemployment  risk,  but  the  difference  with  seniors  is  narrower  than  in  SE.  Job 
mobility  rate  is  high,  although  seniors’  mobility  remains  more  limited.  The  risk  of 
downgrading  for  juniors  remains  present. 


Figure  1 (in  the  annex)  shows  the  cluster  dendrogram  (tree  resulting  from  the  cluster).  Four 
clusters  are  clearly  identifiable  and  seem  to  be  a relevant  grouping  choice. 

(1)  First  cluster:  Germany,  Denmark,  the  Netherlands  and  Austria. 

(2)  Second  cluster:  Spain. 

(3)  Third  cluster:  Italy  and  Greece. 

(4)  Fourth  cluster:  Belgium,  Ireland,  France,  Finland,  Luxembourg,  Portugal,  Sweden  and 
United  Kingdom. 

Figures  2 to  8 summarise,  for  each  cluster,  the  indicators’  values  for  every  individual  used  in 
the  analysis,  that  is  every  country-year.  The  plotted  boxes  represent  the  range  of  median 
observed  values  (between  the  first  and  the  third  quartile)  while  the  thin  stick  represents  the 
whole  range  of  observed  values  within  the  cluster. 


ERIC 


3.15 


248 


Figure  2 : Box  Plot  (Q1,  Q3)  of  the  share  of  trainees  who  have  a job  among  people 
participating  in  a vocational  & post-secondary  course  in  an  institution  belonging  to  the 
education  and  training  system,  by  cluster  - (min  and  max  indicated) 


Figure  3 : Box  Plot  (Q1,  Q3)  of  the  share  of  employed  juniors  who  have  a vocational 
diploma  (ISCED  3 level)  as  highest  diploma,  by  cluster  - (min  and  max  indicated) 


Cluster  1 Cluster  2 Cluster  3 Cluster  4 


Figure  4 : Box  Plot  (Q1 , Q3)  of  the  ratio  of  unemployment  rate  between  juniors  and 
seniors,  by  cluster  - (min  and  max  indicated) 


n m 


Figure  5:  Box  Plot  (Q1t  Q3)  of  the  level  of  juniors'  mobility  out  of  Jobs,  by  cluster  - (min 

and  max  indicated) 


| 60 

i 

; 50 

■ 40 

30 

i 20 

10 

i 

! 0 

Cluster  1 Cluster  2 Cluster  3 Cluster  4 


i 

i 


i 


Figure  6:  Box  Plot  (Q1,  Q3)  of  the  level  of  seniors’  mobility  out  of  jobs, 
by  cluster  • (min  and  max  Indicated) 


ERIC 


Hgure  7 : Box  Plot  (Q1 , Q3)  of  the  ISE  scores  ratio  for  intermediate  and  higher  levels  of 
education,  by  cluster  • (min  and  max  indicated) 


Cluster  1 


Cluster  2 


Cluster  3 


Cluster  4 


H ^ A 


Hgure  8 : Box  Plot  (Q1 , Q3)  of  the  dispersion  of  juniors  in  sectors  (coefficient  of 
variation),  by  cluster  • (min  and  max  indicated) 


In  first  cluster  countries,  vocational  training  is  widespread  and  vocational  trainees  & post- 
secondary students  often  have  a job  while  participating  in  the  IETS,  as  shown  in  figures  2 and 
3.  Unemployment  risk  is  comparable  for  juniors  and  seniors,  the  ratio  is  always  smaller  than 
1 .6.  New  entrants  are  less  often  downgraded  then  in  other  countries,  even  though  there  is  a 
wide  range  of  situations  within  the  cluster.  They  are  not  concentrated  in  specific  industries. 
This  cluster  is  the  closest  to  the  regulated  inclusion  type. 


Cluster  2 is  a single  country,  Spain.  What  makes  Spain  singular  are  the  very  high  mobility 
rates,  for  juniors  as  well  as  for  seniors.  Juniors  are  not  concentrated  in  specific  industries. 
Unemployment  rate  for  juniors  is  less  than  twice  as  high  as  for  seniors.  All  this  suggests  a 
SE+CR  model,  with  high  job  flexibility,  but  with  vocational  training  still  limited. 

Cluster  3 consists  of  two  other  Mediterranean  countries,  Italy  and  Greece.  Here,  education 
and  training  rarely  overlap  with  work  experience  (figure  2).  Job  mobility  is  limited.  Juniors' 
unemployment  rate  is  more  than  three  times  higher  than  that  of  seniors.  Juniors'  jobs  are 
located  in  specific  industries  (as  shown  in  figure  8).  All  this  suggests  that  the  cluster  is  the 
closest  to  the  selective  exclusion  model. 

Cluster  4 is  more  heterogeneous.  The  unemployment  probability  of  juniors,  compared  to 
seniors,  is  higher  than  in  clusters  1 and  2.  But  the  unemployment  ratio  is  far  behind  those  of 
cluster  3 countries.  Job  mobility  is  intermediate,  and  juniors  are  much  more  mobile  than 
seniors.  In  most  of  the  countries  of  this  cluster,  the  downgrading  risk  for  juniors  is  high 


3.18 


(figure  7).  Dual  statuses,  initial  vocational  training  and  mobility  across  industries  for  juniors 
show  more  heterogeneity  within  the  cluster.  On  the  whole,  this  cluster  shows  different 
combinations  from  the  SE+CR  type.  The  main  difference  from  cluster  2 (Spain)  is  less 
developed  job  mobility. 

Heterogeneity  within  clusters 

Actual  systems  are  more  complex  than  ideal-types  of  regulatory  systems.  Cluster  2 and  four 
do  not  closely  fit  one  theoretical  type,  but  appear  to  be  to  different  degrees  intermediate 
between  them.  And  within  each  cluster,  we  can  find  various  dissimilarities.  Within  cluster  1, 
the  proportion  of  ‘initial  trainees’  is  uneven.  Furthermore,  downgrading  of  new  entrants, 
unlike  in  the  regulated  inclusion  model  is  observed  for  certain  countries.  The  share  of  ‘dual 
statuses’  among  trainees,  the  weight  of  initial  vocational  training  or  the  dispersion  index  of 
juniors  across  industries  are  far  from  equal  within  cluster  four. 

In  order  to  illustrate  the  common  patterns  within  clusters,  as  well  to  measure  the 
heterogeneity  they  contain,  we  tested  several  models. 

Model  A:  unemployment  probability 

Model  A is  a logistic  regression  estimating  the  probability  of  being  unemployed  according  to 
time,  gender,  class  of  experience  (junior/senior),  cluster,  experience  by  cluster,  country  within 
cluster(A2,A3),  experience  within  country  within  cluster(A3).  Figure  9 gives  a graphic 
representation  of  the  results  (odds  ratios  by  country  within  clusters,  for  juniors  and  seniors, 
model  in  annex  2). 


Figure  9 : Representation  of  odds  ratio  to  be  unemployed  - model  3 (ref. : Danish  seniors=l) 


ERIC 


3.19 


252 


Cluster  coefficients  for  juniors  show  that,  on  average,  juniors  are  more  often  unemployed. 
Spanish  (cluster  2)  juniors  have  a higher  risk  of  being  unemployed  (with  an  odds  ratio  near 
eight,  compared  to  Danish  seniors),  but,  compared  to  their  seniors,  they  are  not  the  most 
disadvantaged.  Unemployment  risk  for  cluster  3 juniors  is  not  so  high,  but  it  is  higher,  if 
evaluated  relative  to  their  seniors.  Cluster  4 juniors  have  higher  unemployment  probabilities 
than  their  seniors,  but  the  difference  is  far  less  marked  than  in  clusters  2 and  3.  The  less 
disadvantaged  are  by  and  large  cluster  1 juniors. 

National  unemployment  levels  are  very  different,  and  all  country  coefficients  are  significant. 
However,  the  relative  positions  of  juniors  and  seniors  are  fairly  homogeneous  within  clusters: 
German  juniors  are  a bit  less  exposed  to  unemployment  than  the  average  of  the  cluster  (with  a 
95%  confidence  interval).  In  contrast  to  their  seniors,  Belgian,  Irish  and  Finnish  juniors  have 
a lower  additional  risk  of  being  unemployed  than  the  other  juniors  of  cluster  4. 


Model  B:  chances  of  ' job  mobility 

Model  B is  a logistic  regression  estimating  the  probability  of  leaving  the  job  held  one  year 
previously  according  to  time,  gender,  class  of  experience  (junior/senior),  cluster,  experience 
by  cluster,  country  within  cluster(A2,A3),  experience  within  country  within  cluster(A3). 
Figure  10  gives  a graphic  representation  of  the  results  (odds  ratios  by  country  within  clusters, 
for  juniors  and  seniors,  model  in  annex  2). 


Figure  10:  Representation  of  odds  ratio  of  being  mobile  - model  3 (ref. : Danish  seniors=l) 


O 

ERIC 


3.20 


253 


Job  mobility  level  is  low  for  cluster  3 seniors,  medium  for  cluster  1 and  4,  and  high  for  cluster 
2.  Juniors  are  significantly  more  mobile  than  their  seniors. 

Cluster  3 juniors  are  closer  to  their  seniors’  mobility  chances  than  juniors  from  other  clusters. 

Within  cluster  1,  mobility  rates  have  very  different  levels,  the  Danish  being  more  mobile. 
Within  cluster  4,  mobility  is  more  frequent  in  Finland  and  the  UK,  less  frequent  in  Belgium 
and  Luxembourg. 

Relative  to  their  seniors,  German  juniors  have  a lower  additional  risk  than  in  other  countries 
of  cluster  1 . It  is  the  opposite  for  the  Dutch  juniors. 

The  country-specific  risk  of  mobility  for  juniors  is  limited  within  cluster  4:  in  proportion  to 
their  seniors,  only  British  and  Portuguese  juniors  have  a specific  (lower)  risk. 


Model  C:  Occupational  status  by  ISCED  level 

Model  C is  a set  of  two  linear  regressions  estimating  the  level  of  occupational  status 
(measured  by  the  ISEI  score)  according  to  time,  gender,  class  of  experience  (junior/senior), 
cluster,  experience  by  cluster,  country  within  cluster(A2,A3),  experience  within  country 
within  cluster(A3).  The  first  regression  is  devoted  to  people  with  ISCED  3 and  the  second  to 
people  with  ISCED  5 to  7.  Figure  1 1 gives  a graphic  representation  of  the  results  (odds  ratios 
by  country  within  clusters,  for  juniors  and  seniors,  model  in  annex  2). 


Figure  11:  Representation  of  estimates  of  ISEI  values  - models  3.1  & 3.2 


8 DK  (5-7)  a AT  (5-7)  . . a.  . - DE  (5-7)  S N L (5-7)  « ES  (5-7)  - -■ IT  (5-7) 

♦ EL  (5-7)  — « FR  (5-7)  • RU  (5-7)  - -o-  -IE  (5-7)  - - « - - - LU  (5-7)  - -O PT  (5-7) 

• -•□--•  SE  (5-7) ■ FI(5-7)  8 E (5-7)  — * — OK  (3)  ---a..-AT(3)  ° OE  (3) 

— a — NL{3)  ES  (3)  — • — IT  (3)  — ♦—  - EL  (3)  FR  (3)  RU  (3) 

—a IE  (3)  ° LU  (3)  — a — P T (3)  = SE  (3)  — ■ — F I (3)  BE  (3) 


3.21 


254 


A risk  of  downgrading,  measured  by  ISEI  scores,  is  present  for  higher  education  graduates  in 
cluster  3 and  4 countries.  It  is  very  limited  in  cluster  2.  Third  level  juniors  are  less 
disadvantaged  in  cluster  4 countries  than  in  cluster  3.  For  higher  education  graduates,  the  ISEI 
average  score  for  cluster  1 juniors  is  higher  than  for  seniors  (close  to  3 points).  But,  within 
clusters,  results  are  fairly  heterogeneous:  German  juniors  are  more  advantaged  than  Danish 
ones,  Austrian  and  above  all,  Dutch  significantly  less.  The  downgrading  risk  for  Dutch  juniors 
is  even  higher  than  in  cluster  4 countries.  Downgrading  risk  for  Greek  juniors  is  fairly  below 
the  Italian  one.  Within  cluster  4,  the  risk  of  having  a lower  ISEI  score  is  significantly  higher 
in  Ireland,  the  UK  and  Belgium,  lower  in  Portugal. 

At  level  ISCED  3,  a downgrading  risk  for  juniors  is  present  in  all  countries  except  the 
Netherlands,  although  to  a smaller  extent  than  for  ISCED  5-7.  There  is  no  significant 
difference  within  cluster  4.  Within  cluster  1,  Dutch  juniors  have  a significantly  better  position, 
and  German  have  a lower  relative  score  than  Danish  or  Austrian  juniors. 


Model  D:  share  of  juniors'  employment  across  industries 

Model  D is  a linear  regression  estimating  the  absolute  difference  in  the  mean  of  juniors'  share 
in  economic  activities  according  to  time,  cluster,  country  within  cluster  (A2,  A3)  and 
economic  activities  (A3). 

Cluster  1 and  2 show  the  smallest  variation  of  juniors’  employment  across  industries.  Cluster 
4 countries  have  an  intermediate  position,  with  juniors  more  concentrated  in  certain  industries 
than  in  the  former  countries,  but  less  so  than  in  cluster  3.  In  cluster  3 countries,  juniors  are 
more  often  employed  in  specific  sectors. 

Country  variation  is  found  only  within  cluster  4:  Swedish,  Irish,  and  above  all,  Portuguese 
juniors  are  more  often  employed  in  a specific  subset  of  industries. 


Conclusion 


Juniors'  situation  on  the  labour  market  varies  strongly  across  countries.  This  paper  shows  that, 
compared  to  seniors,  their  relative  position  on  the  LM  is  also  distinct.  Empirical  evidence 
brings  out  four  contrasting  sets  of  countries.  According  to  the  assumptions  we  made  (table  1), 
cluster  1 is  the  closest  to  regulated  integration,  cluster  3 to  selective  exclusion,  cluster  2 and  4 


ERiC 


3.22 


255 


are  several  combinations  of  SE+CR  type.  Cluster  2 (Spain)  is  characterised  by  the  highest 
levels  of  unemployment  and  job  mobility. 

Nevertheless,  each  national  system  is  much  more  complex  than  the  polar  cases.  The 
supplementary  risk  of  being  unemployed  attached  to  the  junior  condition  is  fairly 
homogeneous  within  the  four  sets.  But  the  national  systems  are  far  from  equivalent  : 
downgrading  risks  for  juniors  are  high  in  the  Netherlands,  unlike  in  the  typical  RI  model.  The 
SE+CR  model  is  supposed  to  provide  more  opportunities  for  new  entrants,  but  in  Ireland, 
Sweden  & Portugal,  juniors  are  still  concentrated  in  certain  industries,  like  in  a pure  SE 
model.  It  is  supposed  to  favour  vocational  education,  but  it  is  less  the  case  in  Spain,  Belgium 
or  Portugal  than  in  other  countries  of  clusters  2 & 4.  Mobility  chances  also  show  a certain 
heterogeneity  within  clusters. 

Some  comments  can  be  made  about  the  general  framework  we  used.  First,  public  policy 
issues  are  not  explicitly  taken  into  account.  It  is  particularly  problematic  in  the  countries 
where  the  real  system  is  near  the  SE  or  SE+CR  models.  In  the  past  twenty  years,  most  of  these 
countries  have  developed  education,  training  and  labour  market  policies  which  attempt  to 
prevent  youth  exclusion.  Second,  the  dynamic  of  national  systems  is  imperfectly  captured. 

In  relation  to  public  policy  issues,  Gautie  and  Lefresne  (1997)  adapted  the  Garonna  and  Ryan 
model  in  order  to  define  four  youth  inclusion  patterns,  incorporating  public  policies: 
“insertion  negociee”  (negotiated  inclusion)  - same  as  regulated  inclusion-,  “insertion 
reglementee”  (state  regulated  inclusion)  -such  as  quota  policy  for  disabled  workers,  or  other 
disadvantaged  minorities-,  “insertion  incitee”  (public  incentives  for  youth  inclusion)  -in  a 
former  selective  exclusion  model,  policy  makers  give  incentives  to  firms  for  hiring  new 
entrants  by  creating  specific  low-wage  contracts,  or  by  giving  extra  training  to  the  young 
unemployed-  and  “insertion  concurrentielle”  (competitive  inclusion)  -in  a former  selective 
exclusion,  or  regulated  inclusion  model,  public  policies  try  to  promote  competition,  by 
suppressing  controls  on  hiring  and  redundancies,  or  by  making  job  contracts  regulation  more 
flexible.  In  the  latter  case,  an  evolution  towards  the  competitive  regulation  model  is  an 
explicit  goal  for  public  policy. 

This  adapted  model  could  be  taken  into  account,  with  quantitative  information  on  public 
policy.  More  generally,  other  actors'  strategies  could  be  integrated  with  more  available 
information. 


O 

ERIC 


3.23 


256 


In  relation  to  the  dynamic  of  national  systems,  many  studies  have  shown  that  each  actual 
system  is  submitted  to  endogenous  or  exogenous  shocks.  Increasing  international  competition, 
biased  technical  progress,  new  regulations  of  competition  in  post-fordist  economies  affect  the 
demand  for  competencies,  as  well  as  the  nature  of  the  relationship  between  employers  and 
employees  (Petit  1998).  These  shocks  have  consequences  for  working  life,  career  progression 
rules,  etc.  For  instance,  it  is  well  known  that  in  France  career  prospects  for  workers  are  not  the 
same  today  as  they  were  for  the  former  generation,  because  of  the  competition  between 
qualified  new  entrants  and  experienced  workers  without  qualifications  (Goux  1991). 
Buechtemann  and  Verdier  (1998)  quote  the  same  phenomenon  in  Germany:  promotion  from 
skilled  worker  to  Meister  or  Techniker  tends  to  be  less  frequent. 

With  longer  time  series  and  more  microdata,  testing  how  the  national  systems  are  affected  by 
these  shocks  will  be  possible. 


3.24 


2 57 


References 


Arrow  K.  J.,  1993  : “Higher  Education  as  a filter”,  Journal  of  Public  Economics,  2.. 

Ashton  D.,  1994  : “Understanding  changes  in  youth  labour  markets:  a conceptual 
framework”,  In  The  determinants  of  transitions  in  youth.  Papers  from  the  conference 
organised  by  the  ESF  Network  on  Transitions  in  Youth,  CEDEFOP  and  GRET 
(Universitat  autonoma  de  Barcelona).  Barcelona  20-21  September  1993  CEDEFOP, 
Berlin,  1994,  329  p. 

Becker  G.,  1993  : Human  capital.  A theoretical  and  empirical  analysis,  ( with  special 
reference  to  education).  Chicago/  London,  The  University  of  Chicago  Press,  (3e  ed.). 

Boyer  R.  and  M.  Juillard,  1998:  “Mondialisation  et  construction  europeenne  ; Le  rapport 
salarial  et  les  ajustements  du  marche  du  travail  tendent-ils  a converger?”,  Rapport  pour 
le  Commissariat  General  du  Plan. 

Buechtemann  C.  and  E.  Verdier,  1998  : “Education  and  training  regimes,  macro-institutional 
evidence”.  Revue  dEconomie  Politique,  mai-juin,  n°  3,  pp.  291-320. 

Cahuc  P.  and  A.  Zylberberg,  1996 : Economic  du  travail.  De  Boeck  & Larcier  s.a., 
Departement  De  Boeck  Universite,  Bruxelles. 

Doeringer  P.  and  M.  Piore,  1971  : Internal  labour  markets  and  manpower  analysis, 
Lexington  : Heath  Lexington  Books. 

Escofier  B.  and  J.  Pages,  1988  : Analyses  factorielles  simples  et  multiples.  Objectifs, 
methodes  et  interpretation.  Paris:  Dunod/Bordas. 

EUROSTAT,  1996  : L’enquete  communautaire  sur  les  Forces  de  Travail.  Methodes  et 
definitions.  Luxembourg:  Office  des  publications  officielles  des  Communautes 
europeennes. 

Ganzeboom  H.,  and  D.  Treiman,  1996:  “Internationally  Comparable  Measures  of 
Occupational  Status  for  the  1988  International  Standard  Classification  of 
Occupations”.  Social  Science  Research  25,  201-239. 

Garonna,  P.  and  P.  Ryan,  1989  : “Le  travail  des  jeunes,  les  relations  professionnelles  et  les 
politiques  sociales  dans  les  economies  avancees”,  Formation  - Emploi  n°25,  janvier  - 
mars. 

Garonna,  P.  and  P.  Ryan,  1991  : “The  regulation  and  deregulation  of  youth  economic 
activity”,  in  The  problem  of  youth,  Ryan,  Garonna  and  Edwards  (eds.),  Houndmills: 
Macmillan. 

Gautie,  J.  et  F.  Lefresne,  1997  : “La  politique  de  l’emploi  et  sa  representation  de  l’entreprise”, 
in  La  Revue  de  VIRES,  n°23,  hiver. 

Gollan  J.,  1937  : Youth  in  British  industry.  Gollancz,  London. 

Goux,  D.,  1991  : “Coup  de  frein  sur  les  carrieres”,  in  Economic  et  Statistique,  n°249, 
decembre. 


253 


Lindbeck  A.  and  D.  J.  Snower,  1988  : The  insider  outsider  theory  of  employment  and 
unemployment,  the  M.  I.  T.  press,  Cambridge. 

Marsden,  D.,  1986  : The  end  of  economic  man  ? Custom  and  competition  in  labour  markets. 
Brighton  (RU),  Wheatsheaf  Books  Ltd. 

Marsden  D.,  1991  : “Institutions  and  labour  mobility  : occupational  and  internal  labour 
markets  in  Britain,  France,  Italy  and  Germany”  in  Labour  relations  and  economic 
performance  R.  Brunetta  and  C.  dell’ Aringa  (eds.),  Houndmills:  Macmillan. 

Marsden  D.,  1993  : “Trade  Union  action  and  labour  market  structure”,  in  International 
Integration  and  Labour  Market  Organisation,  P.A.  Castro,  P.  Mehaut,  Jill  Rubery 
(eds.).  Academic  Press. 

Marsden  D.,  1992  : “Segmentation  des  marches  du  travail  et  chomage  de  longue  duree”,  in  Le 
chomage  de  longue  duree.  Comprendre,  agir,  evaluer,  P.  Bouillaguet  et  C.  Guitton 
(eds.).  Paris:  Syros  Alternatives. 

Marsden  D.  and  P.  Ryan,  1986  : “Where  do  young  workers  work  ? Youth  employment  by 
industry  in  various  European  economies”,  British  Journal  of  Industrial  Relations , vol. 
XXIV,  n°l. 

Marsden  D.  and  P.  Ryan,  1990  : “Institutional  aspects  of  youth  employment  and  training 
policy  in  Britain”,  British  Journal  of  Industrial  Relations , vol.  XXVIII,  n°3. 

Marsden  D.  and  P.  Ryan,  1991  : “The  structuring  of  youth  pay  and  employment  in  six 
European  economies”,  in  The  problem  of  youth,  Ryan,  Garonna  and  Edwards  (eds.), 
Houndmills:  Macmillan. 

Marsden  D.  and  J.-F.  Germe,  1991  : “Young  people  and  entry  paths  to  long-term  jobs  in 
France  and  Great  Britain”,  in  The  problem  of  youth,  Ryan,  Garonna  and  Edwards 
(eds.),  Houndmills:  Macmillan. 

Maurice  M.,  F.  Sellier  et  J.-J.  Sylvestre,  1982  : Politique  d ’education  et  organisation 
industrielle  en  France  et  en  Allemagne,  Paris:  PUF. 

Mehaut  P.,  1993  : “Further  education,  vocational  training  and  the  labour  market : the  French 
and  German  systems  compared”,  in  International  Integration  and  Labour  Market 
Organisation,  ;PA.  Castro,  P.  Mehaut,  Jill  Rubery  (eds.).  Academic  Press. 

Muller,  W.  and  Y.  Shavit,  1998  : From  School  to  Work.  A comparative  study  of  educational 
qualifications  and  occupational  destinations.  Oxford:  Clarendon  Press. 

OCDE,  1998  : Regards  sur  l ’education.  Paris:  Editions  de  l’OCDE. 

Petit  P.,  1998  : Travail  et  emploi  dans  une  economie  fondee  sur  le  savoir.  Communication  au 
colloque  organise  en  novembre  1998  a Vancouver  par  la  Commission  Europeenne 
(DG  V)  et  l’Universite  de  British  Columbia,  mimeo. 

Ryan  P.,  1991  : “Trade  Unions  policies  towards  the  Youth  Training  Scheme  in  Great  Britain  : 
the  arguments”,  in  The  problem  of  youth,  Ryan,  Garonna  and  Edwards  (eds.), 
Houndmills:  Macmillan. 


O 

ERIC 


326  259 


Ryan  P.,  1999  : “The  School-to-Work  Transition  : a cross-national  perspective”,  mimeo, 
University  of  Cambridge. 

Spence  M.,  1974  : Market  Signalling;  Informational  Transfer  in  Hiring  and  Related 
Screening  Processes,  Harvard  U Press,  Cambridge. 

Thurow  L.,  1975  : Generating  inequality,  Macmillan,  London. 

Verdier  E.  and  M.  Mobus,  1999  : “Diplomes  professionnels  et  coordination  de  la  formation  et 
de  Temploi : Elaboration  d’un  signal  en  France  et  d’une  regie  en  Allemagne”,  Revue 
d’economie  Publique,  forthcoming. 


ERIC 


3.27 


2G0 


Annex  1 ; Figure  1 


Dendrogram  of  the  cluster  analysis 


93DE 
94  D3 
95DB 
96  DE 
97DB 

96  DK 
95DK 
97DK 
94DK 

94  NIi 
93NL 

95  NL 

97  NL 

96  NL 
93DK 
95AT 
96AT 
97AT 
97ES 
96ES 
95ES 
94  ES 
93ES 
94  IT 

93  IT 

97  EL 
96  EL 
95EL 

94  EL 

93  EL 
96IT 
95IT 
97IT 
9 7 UX 
96UK 
95UK 
93UK 

94  UX 
94  FR 

93  PH 

96  FR 
95FR 

97  FR 
97FI 
97SB 
97LU 

96  LU 
9SLU 

94  LU 
9 3 LU 

97  PT 
94  PT 
95PT 

96  PT 

93  PT 
95IE 

94  IE 

97  IE 
96  IE 

93  IE 

94  BE 
95BB 

96  BE 

97  BE 
93BB 


3. 


Annex  2 : models 


Results  of  logistic  regression  analysis  of  unemployment 


O 

ERIC 


Model 

1 

2 

3 

Constant 

-2,64  " 

-2,71  ** 

-2,76  ** 

Tima  trend  (1993=0) 

0,00 

0,00 

0,00 

Gender 

Males 

Ref. 

Ref. 

Ref. 

Females 

0,29  ** 

0,29  ** 

0,29  ** 

LM  experience 

Juniors 

0,28  " 

0,28  ** 

0,45  ** 

Seniors 

Ref. 

Ref. 

Ref. 

Cluster  solution 

Ouster  1 

Ref. 

Ref. 

Ref. 

Ouster  2 

1,14  ** 

1,22  ** 

1,26  ** 

Ouster  3 

0,19  ** 

0,30  ** 

0,34  ** 

Ouster  4 

0,21  ** 

0,50  ** 

0,54  ** 

Cluster  solution  * experience 

Ouster  2 * Juniors 

0,55  " 

0,54  " 

0,38  ** 

Ouster  3 * Juniors 

1,25  ~ 

1,25  ** 

1,08  ** 

Ouster  4 * Juniors 

0,63  ** 

0,64  *• 

0,52  ** 

Cluster  solution  * country 

Ouster  1 * Denmark 

Ref. 

Ref. 

Ouster  1 * Austria 

-0,45  ** 

-0,44  ** 

Ouster  1 * Germany 

0,15  ** 

0,20  " 

Ouster  1 * Netherlands 

-0,17  ** 

-0,16  ** 

Ouster  3 * Italy 

Ref. 

Ref. 

Ouster  3 * Greece 

-0,23  ** 

-0,23  ** 

Ouster  2 * Spain 

Ref. 

Ref. 

Ouster  4 * Luxembourg 

-1,46  ** 

-1,43  ** 

Ouster  4 * Fbrtugal 

-0,66  " 

-0,66  ** 

Ouster  4 * France 

Ref. 

Ref. 

Ouster  4 * United  Kingdom 

-0,39  " 

-0,38  " 

Ouster  4 * Ireland 

0,01 

0,16  ** 

Ouster  4 * Sw  eden 

-0,25  ** 

-0,28  ** 

Ouster  4 * Finland 

0,07 

0,13  ** 

Ouster  4 * Belgium 

-0,34  ** 

-0,30  ** 

Cluster  solution  * country  * LM  experience 

Ouster  1 * Denmark  * Juniors 

Ref. 

Ouster  1 * Austria  * Juniors 

0,01 

Ouster  1 * Germany  * Juniors 

-0,21  * 

Ouster  1 * Netherlands  * Juniors 

0,01 

Ouster  3 * Italy  * Juniors 

Ref. 

Ouster  3 * Greece  * Juniors 

0,01 

Ouster  2 * Spain  * Juniors 

Ref. 

Ouster  4 * Luxembourg  * Juniors 

-0,12 

Ouster  4 * Fbrtugal  * Juniors 

-0,02 

Ouster  4 * France  * Juniors 

Ref. 

Ouster  4 * United  Kingdom  * Juniors 

-0,04 

Ouster  4 * Ireland  * Juniors 

-0,56  ** 

Ouster  4 * Sw  eden  * Juniors 

0,13 

Ouster  4 * Finland  * Juniors 

-0,30  ** 

Ouster  4 * Belgium  * Juniors 

-0,18  ** 

Model  Chi2 

20769  ** 

22334  ** 

22407  ** 

DF 

9 

20 

31 

N 

624185 

624185 

624185 

* = p < 0.05;  - = p < 0.01 
Ref.  = reference  category 


26; 


3.29 


Results  of  logistic  regression  analysis  of  job  mobility 


Model 

1 

2 

3 

Constant 

-2,07  ** 

-1,75  ** 

-1,79  ** 

Time  trend  (1993=0) 

-0,01  ** 

-0,01  ** 

• 

o 

o 

3 

Gender 

Males 

Ref. 

Ref. 

Ref. 

Females 

0,14  ** 

0,14  ** 

0,13  ** 

LM  experience 

Juniors 

0,96  ** 

0,92  " 

1,08  " 

Seniors 

Ref. 

Ref. 

Ref. 

Cluster  solution 

Ouster  1 

Ref. 

Ref. 

Ref. 

Ouster  2 

0,91  ** 

0,58  ** 

0,62  " 

Ouster  3 

-0,44  ** 

-0,78  ** 

-0,73  ** 

Ouster  4 

0,12  ** 

-0,33  ** 

-0,30  ** 

Cluster  solution  * experience 

Ouster  2 * Juniors 

0,36  ** 

0,41  ** 

0,25  ** 

Ouster  3 * Juniors 

-0,07 

-0,03 

• 

o 

To 

3 

Ouster  4 * Juniors 

0,22  ** 

0,25  ** 

0,17  * 

Cluster  solution  * country 

Ouster  1 * Denmark 

Ref. 

Ref. 

Ouster  1 * Austria 

-0,70  ** 

-0,67  ** 

Ouster  1 * Germany 

-0,30  ** 

-0,25  ** 

Ouster  1 * Netherlands 

-0,41  ** 

-0,42  ** 

Ouster  2 * Spain 

Ref. 

Ref. 

Ouster  3 * Italy 

Ref. 

Ref. 

Ouster  3 * Greece 

0,05 

0,03 

Ouster  4 * France 

Ref. 

Ref. 

Ouster  4 * United  Kingdom 

0,30  ** 

0,32  ** 

Ouster  4 * Ireland 

-0,03 

0,00 

Ouster  4 * Luxembourg 

-0,43  ** 

-0,39  ** 

Ouster  4 * Fbrtugal 

0,02 

0,06  * 

Ouster  4*  Sweden 

0,08 

0,04 

Ouster  4 * Finland 

0,24  ** 

0,23  ** 

Ouster  4 * Belgium 

-0,24  ** 

-0,26  ** 

Cluster  solution  * country  * LM  experience 

Ouster  1 * Denmark  * Juniors 

Ref. 

Ouster  1 * Austria  * Juniors 

-0,10 

Ouster  1 * Germany  * Juniors 

-0,23  ** 

Ouster  1 * Netherlands  * Juniors 

0,31  ** 

Ouster  2 * Spain  * Juniors 

Ref. 

Ouster  3 * Italy  * Juniors 

Ref. 

Ouster  3 * Greece  * Juniors 

0,15 

Ouster  4 * France  * Juniors 

Ref. 

Ouster  4 * United  Kingdom  * Juniors 

-0,14  ** 

Ouster  4 * Ireland  * Juniors 

-0,16 

Ouster  4 * Luxembourg  * Juniors 

-0,21 

Ouster  4 * Fbrtugal  * Juniors 

-0,26  ** 

Ouster  4 * Sw  eden  * Juniors 

0,18 

Ouster  4 * Finland  * Juniors 

0,06 

Ouster  4 * Belgium  * Juniors 

0,08 

Model  Chi2 

19121  ** 

19847,4  ** 

20338  ** 

DF 

9 

20 

31 

N 

524536 

524536 

524536 

* = p < 0.05;  **  = p < 0.01 
Ref.  = reference  category 


3.30  O £ ry 

S-  ? J O 


Results  of  partial  linear  regression  analysis  of  ISEI  occupational  status  by  ISCED  level 


Model 

1. 

1.1 

ISCHD  = 5to  7 

1.2 

1.3 

2.  ISC  HD  = 3 

2.1  2.2 

2.3 

Constant 

57,89  ** 

57,29  ** 

57,36  ** 

39,60  **  38,30  ** 

38,58  ** 

Time  trend  (1993=0) 

-0,07  ** 

-0,09  ** 

-0,09  ** 

-0,10  **  -0,07  ** 

-0,06  ** 

Gender 

Males 

Ref. 

Ref. 

Ref. 

Ref.  Ref. 

Ref. 

Females 

-3,22  ** 

-3,27  ** 

-3,26  ** 

2,35  **  2,36  ** 

2,37  ** 

LM  experience 

Juniors 

3,16  ** 

3,26  ** 

2,91  ** 

-1,14  **  -1,12  ** 

-2,48  ** 

Seniors 

Ref. 

Ref. 

Ref. 

Ref.  Ref. 

Ref. 

Cluster  solution 

Ouster  1 

Ref. 

Ref. 

Ref. 

Ref.  Ref. 

Ref. 

Ouster  2 

-0,78  ** 

-0,10 

-0,18 

2,12  **  3,32  ** 

3,03  ** 

Ouster  3 

9,57  ** 

11,16  ** 

11,18  ** 

5,37  **  7,13  ** 

6,87  ** 

Ouster  4 

2,69  ** 

2,36  ** 

t 

O) 

CO 

1,89  **  1,77  ** 

: 

o 

to 

Cluster  solution  * experience 

Ouster  2 * Juniors 

-3,91  ** 

-4,01  ** 

-3,66  ** 

-4,58  **  -4,60  ** 

-3,24  ** 

Ouster  3 * Juniors 

-8,95  ** 

-8,54  ** 

-9,37  ** 

-2,69  **  -2,54  ** 

-1,41  ** 

Ouster  4 * Juniors 

-6,34  ** 

-6,32  ** 

-4,07  ** 

-2,82  **  -3,09  ** 

-1,84  ** 

Cluster  solution  * country 

Ouster  1 * Denmark 

Ref. 

Ref. 

Ref. 

Ref. 

Ouster  1 * Austria 

10,67  ** 

11,15  ** 

1,77  ** 

1,71  ** 

Ouster  1 * Germany 

-0,32 

-0,65  * 

0,73  ** 

0,32 

Ouster  1 * Netherlands 

4,43  ** 

5,41  ** 

3,89  ** 

4,10  ** 

Ouster  2 * Spain 

Ref. 

Ref. 

Ref. 

Ref. 

Ouster  3 * Italy 

Ref. 

Ref. 

Ref. 

Ref. 

Ouster  3 * Greece 

i 

CM 

CO 

CO 

1 

-4,02  ** 

-3,89  ** 

-4,10  ** 

Ouster  4 * France 

Ref. 

Ref. 

Ref. 

Ref. 

Ouster  4 * United  Kingdom 

1,56  ** 

2,27  ** 

3,15  ** 

3,12  ** 

Ouster  4 * Ireland 

0,60 

1,59  ** 

2,49  ** 

2,44  ** 

Ouster  4 * Luxennbourg 

1,83  * 

2,34  * 

3,16  ** 

2,94  ** 

Ouster  4 * Fbrtugal 

4,15  ** 

3,95  ** 

6,32  ** 

6,38  ** 

Ouster  4 * Sweden 

-1,13  ** 

-1,08  * 

-2,36  ** 

-2,54  ** 

Ouster  4 * Finland 

1,72  ** 

1,26  * 

-0,68 

-0,71 

Ouster  4 * Belgium 

1,20  ** 

1,61  ** 

0,27 

0,32  * 

Cluster  solution  * country  * LM  experience 

Ouster  1 * Denmark  * Juniors 

Ref. 

Ref. 

Ouster  1 * Austria  * Juniors 

-3,42  ** 

-0,17 

Ouster  1 * Germany  * Juniors 

1,72  ** 

-2,10  ** 

Ouster  1 * Netherlands  * Juniors 

-6,00  ** 

2,25  ** 

Ouster  2 * Spain  * Juniors 

Ref. 

Ref. 

Ouster  3 * Italy  * Juniors 

Ref. 

Ref. 

Ouster  3 * Greece  * Juniors 

3,18  ** 

1,27  ** 

Ouster  4 * France  * Juniors 

Ref. 

Ref. 

Ouster  4 * United  Kingdom  * Juniors 

-4,21  ** 

0,25 

Ouster  4 * freland  * Juniors 

-4,22  ** 

0,28 

Ouster  4 * Luxerrbourg  * Juniors 

-2,60 

1,60 

Ouster  4 * Fbrtugal  * Juniors 

1,54  * 

-0,51 

Ouster  4 * Sw  eden  * Juniors 

1,63 

1,34 

Ouster  4 * Finland  * Juniors 

2,63 

0,25 

Ouster  4 * Belgium  * Juniors 

■ 

ro 

o 

00 

J 

-0,45 

SS 

1141015 

1417032 

1497547 

1105039  1651793 

1604776 

F value 

1017,0  ** 

573,8 

392,4  ** 

1178,7  **  748,9  ** 

493,3  ** 

DF 

9 

20 

31 

9 20 

31 

R? 

0,040 

0,05 

0,053 

0,032  0,044 

0,045 

N 

219434 

219434 

219434 

327790  327790 

327790 

* = p < 0.05;  - = p < 0.01 
rf . = reference  category 


Results  of  linear  regression  analysis  of  absolute  variation 
to  the  mean  of  juniors  share  in  economic  activities 


Model 

1 

2 

3 

Constant 

14,99  " 

14,20  ** 

8,33  " 

Time  trend  (1993=0) 

-0,07 

-0,10 

-0,09 

Cluster  solution 

Ouster  1 

Ref. 

Ref. 

Ref. 

Ouster  2 

3,67 

4,55 

3,17 

Ouster  3 

16,10  « 

17,83  ** 

16,93  ** 

Ouster  4 

9,67  ** 

6,49  * 

5,58  * 

Cluster  solution  * country 

Ouster  1 * Denmark 

Ref. 

Ref. 

Ouster  1 * Austria 

4,05 

2,95 

Ouster  1 * Germany 

-1,70 

-1,67 

Ouster  1 * Netherlands 

2,48 

1,72 

Ouster  2 * Spain 

Ref. 

Ref. 

Ouster  3 * Italy 

Ref. 

Ref. 

Ouster  3 * Greece 

-1,69 

-4,45 

Ouster  4 * France 

Ref. 

Ref. 

Ouster  4 * United  Kingdom 

1,36 

1,96 

Ouster  4 * Ireland 

10,47  " 

9,76  ** 

Ouster  4 * Luxembourg 

6,04  * 

5,22 

Ouster  4 * Portugal 

12,49  ** 

11,95  *• 

Ouster  4 * Sweden 

5,74 

7,06  * 

Ouster  4 * Finland 

0,61 

0,93 

Ouster  4 * Belgium 

-4,84 

-4,57 

Economic  activities 

Industry 

Ref. 

Agriculture 

20,43  ** 

Construction 

-0,42 

Wholesale  & Retail  Trade 

5,56  ** 

Hotels  & Restaurants 

21,56  ** 

Transport  & communication 

13,35  ** 

Rnancial  intermediaries 

7,55  ** 

Business  activities 

14,14  ** 

Rjblic  administration 

16,90  ** 

Education 

9,63  ** 

Heatth 

-1,44 

Other  service  activities 

-3,05 

SS 

23960 

38681 

89220 

F value 

20,97  " 

9,50  " 

16,19  " 

DF 

4 

15 

26 

R2 

0,093 

0,149 

0,344 

N 

828 

828 

828 

Dependent  variable  : Y,,  = |Si(  - S4|  / S , w here 

Sjj  is  the  share  of  junior  in  the  Ith  economic  activity  of  the  j01  country 
Sj  is  the  average  share  of  junior  in  ail  the  economic  activities  of  the  j*  country 
* = p < 0.05;  **  = p < 0.01 
Ref.  = reference  category 


3.32 


26  5 


Annex  3 : Average  values  of  indicators  of  the  cluster  analysis 


Tf  00  CO  CD 
N-  LO  Tf 


p|  in  h-  p n- 

o'  oo"  cm"  CO 
oo  in  oo 


CM 

LU 

CO 


CO 

QQ 


CD 

o 


a: 

0 

1 

o 


> 

o 


tj-  in 


in 


|s.  t- co 

|sJCM  00 

o o 00 


OOCOr 
hJ  00  O 00 


O)  CD  CM  00 
Is-"  o"  cm"  in 

T-  00  CM  CM 


p Tt  tt  cq 

Is-"  Is-"  co"  00 
in  Tf  cm  io 


00  cop 

in  Is-"  co"  o" 

CM  T-  T-  CM 


(0 

"O 

i-  c <0 
co  {5  co 

| ei! 


(0 


_ £ 
_ 53  0)  o 
<QO" 


CM 


C 

’co 

Qt 

CO 


CD  t- 

o"  m 


O)  co 
oo"  cm" 


p 

cm"  Tf 

CD  CD 


Is-  00 

in"  co 


oo  in 

xf  h-" 


O CM 

cm"  Is- 
m Is- 
00  00 


oo  in 
o"  co 

Tf  oo 


<D 

O 

?<5 


O)o  00  o CO  O CD 
£2"  S o g>  in  o t£  ^ 

CM  CM  IO  Tf 


OCON-o^^^CM 

C^CM-cD^pCM^ 
T-  Tf  T-  T"  CM  CM 


CD 


CO  O t-  O Is-  CD 


T-T-o^inmcoo 

CDCDCDCDCDCDCDCD 


°l  °°L  ^ ^ ^ °- 

cd"  oo  cd"  co"  o oo"  Is-"  cm’ 


in  t-  oo  t-  oo  ° oo  oo 
cd"  oo"  co"  oo"  in"  h-"  Tf"  cd 

CMOOCMt-CMCMOOCM 


oqcooqo^-oocM 

co"  co"  oo"  co"  cd"  Is-"  in" 

CMCMTtOOOOOOinO 

CMCMt-CMCMCMt-CM 


t-w cm^ p co  t-  pro  cd 

Is-"  Is-"  in"  o"  o"  oo"  in"  cm 
CMCMOOOOTfOOCMCM 


E | 

si5?!  ||?  I 

Li  ^ i n rnn  m 


- ^ ^ 
CO  ' — 


co 

n 

© 

e 

o 


© 

£ 

*♦— 

o 

c 

o 

♦3 

(0 

•C 

5 


c w 

■ I 


II 


© 
O D 
CL 

E 


Q. 

S E 
© © 
e 

o 
c 

3 


© 
© 

03 

c 
o 
E 
© 

e 

o 
'c 
© 
CO 

© © 


03 

C 

o 

E 

© 

2 

o 

c 

3 


0)OO 

5 e E 

.§  o o 
o 2 £ 

*3  © © 

<5  x:  £ 
QC  CO  CO 


© 


? I ^ ^ 


© © 
© > 
<j) 

© c 
c o 

° T3 

s s. 

>»  o 

O Q. 


£ 

o 

c 

© 

© 

"O 

c 

© 

e 

o 

*c 

3^ 

o 

c 

o 

p 

© 

3. 

© 

£ 


5 g 

© E 
© w 

a 5 

© 


Q 

LU 


© O 
© £ 
3 "O 

O c 
© © 


CO 

Q 


I S 

© CO 

1 i 

B 2 

2 a, 

1 1 

N*  5 

3 o 

« a 

2 e 

o © 
<2  5 

3 f 


^ 2 
LU  © 

co  £ 

e “ 

1 ^ 

© .S> 
© © 

0 5 

S3 

2 S 
12  © 

1 3 
•a  s 

o © 

.9  © 

2 2 

© © 

? | 
© P 


I 

© 

T5 

© 

T3 

© 

> 

* Z 
1 ' © 
© © 
12  -c 

3 .2> 


N.  © 
O © 

| 

Q CL 
LU  Vq 

0 5 

CO  g 

c £ 
© *5 
o>  g 

1 % 

1 5 

© (o 
© -C 

S o 


S 12 

01  .2 

it 

a i 
© >> 
© O 

1 1 

2 | 
03  O 

I 2 

1 5 

5 co 


H OC  ->  CO  •• 

z o CQ  CD  CN  O 1 

> x O O uj  £ p 

O O 5 2 52  iZ  > 


CD 

CvJ 


CD 

CD 

C\» 


o 

ERLC 


3.33 


January  2001 


Prepared  as  part  of  the  TSER  project: 

Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe 


European  Perspectives  on  Labour  Market  Entry: 
A Matter  of  Institutional  Linkages  between 
Training  Systems  and  Labour  Markets? 


Markus  Gangl 


Mannheim  Centre  for  European  Social  Research  (MZES) 

University  of  Mannheim 

P.O.  Box  10  34  62 

D-68131  Mannheim 

Germany 

E-mail:  Markus.Gangl@mzes.uni-mannheim.de 


263 


WORKING  PAPERS 


Abstract 


The  nature  of  the  linkage  of  education  and  training  systems  to  the  labour  market  is  often  claimed  to 
crucially  affect  labour  market  integration  in  modern  economies.  More  specifically,  most  current 
comparative  research  assumes  a more  strongly  qualification-based  allocation  in  training  systems 
allowing  for  early  occupational  specialization  as  compared  to  more  experience-based  allocation 
mechanisms  where  such  arrangements  are  absent.  Building  on  this  basic  idea,  the  paper  develops  a 
set  of  institutional  predictions  about  consequences  for  patterns  of  labour  market  entry  in  these 
systems.  This  framework  is  then  applied  in  exploratory  analyses  for  twelve  member  states  of  the 
European  Union.  From  these,  three  distinct  patterns  of  early  labour  market  experiences  empirically 
emerge:  first,  a non-experience  based  pattern  for  those  continental  European  countries  with  extensive 
vocational  training  systems,  second,  a strongly  experience-based  allocation  pattern  in  those  Northern 
European  countries  lacking  such  systems,  and,  finally,  a particular  and  theoretically  unexpected 
pattern  among  the  group  of  Southern  European  countries.  While  the  first  contrast  appears  broadly 
consistent  with  current  institutionalist  arguments  about  the  impact  of  interlinked  training  systems  and 
labour  markets,  the  explanation  for  the  peculiarity  of  Southern  Europe  needs  both  further  investigation 
and  additional  conceptual  tools. 


1.  Structured  Entries:  the  importance  of  Institutional  Contexts 

Broadly  speaking,  the  experiences  of  young  people  at  labour  market  entry  depend  on  the  resources 
they  bring  into  the  market,  the  evaluation  of  these  within  the  labour  market  and  their  transformation 
into  the  attainment  of  labour  market  positions.  As  uncontroversial  as  this  is,  social  science  research 
has  long  been  focused  both  on  establishing  a theoretical  understanding  of  this  attainment  process  and 
on  arriving  at  empirical  estimates  of  the  attainment  function.  Be  it  economic  research  in  the  tradition  of 
human  capital  models  or  be  it  sociological  status  attainment  research,  the  main  interest  generally  has 
been  the  role  of  individual  resources  - namely  educational  qualifications  and  work  experience  - and 
ascriptive  factors  like  gender,  social  origin  or  ethnicity  in  the  attainment  process.  As  such,  inequalities 
between  individuals  have  been  highlighted  and  theoretical  understanding  of  attainment  is  mainly 
driven  by  their  determinants  - at  the  expense  of  broader  institutional  or  structural  factors. 

With  the  proliferation  of  comparative  research,  differences  in  attainment  processes  between  countries 
have  been  constantly  and  reliably  established.  Specifically  in  the  study  of  labour  market  entry  and 
transitions  from  school  to  work,  major  recent  contributions  have  concluded  that  individual  resources, 
namely  education,  clearly  matter  for  labour  market  attainment  in  all  industrial  societies,  yet  at  the 
same  time,  the  way  they  do  so  varies  between  countries  (cf.  the  review  in  Kerckhoff  1995;  Muller  and 
Shavit  1998;  Shavit  and  Muller,  forthcoming;  Kerckhoff  1996,  forthcoming;  Hannan  et  al.  1997;  Brauns 
et  al.  1998;  Rosenbaum  et  al.  1990;  Allmendinger  1989;  other  related  studies  comprise  work  by 
Ashton  1988,  1994,  1997;  Ashton  et  al.  1990;  Sengenberger  1987).  These  findings  indeed  add  to 
previous  research  at  the  national  level  as  the  similarities  and  differences  between  societies  draw 
attention  to  the  embeddedness  of  attainment  processes  within  specific  institutional  contexts,  thus 
fostering  and  necessitating  institutional  explanations  in  comparative  research. 

Actually,  much  of  current  comparative  research  is  based  on  the  argument  of  a dichotomy  of 
stratification  systems,  each  representing  a specific  type  of  linkage  between  the  institutional  structure  of 
education  and  training  systems  and  the  stratification  of  labour  market  careers  (cf.  Kerckhoff  1995; 
Muller  and  Shavit  1998;  Hannan  et  al.  1999):  on  the  one  hand,  there  is  one  type  of  system,  regularly 
claimed  to  operate  in  Austria  or  Germany,  tied  to  a horizontally  differentiated  education  and  training 
system  providing  highly  specific,  occupationally  relevant  initial  training,  exhibiting  high  employer 
involvement  in  training  provision  and  strong  occupational  entry  labour  markets.  The  second  type  of 
system  is  more  of  the  French  or  U.S.  type  where  initial  education  is  largely  school-based  and  de- 
coupled from  the  labour  market,  being  in  consequence  more  general  in  nature  and  less  tailored  to  the 
youth’s  specific  future  work  tasks.  Effectively,  as  will  be  argued  in  more  detail  below,  this  institutional 
argument  posits  the  existence  of  different  institutional  equilibria  in  terms  of  the  relative  reliance  of 
market  matching  processes  on  either  educational  certification  or  labour  force  experience  and  mobility. 
Much  of  institutional  theorizing  in  comparative  reseach  on  labour  markets  and  social  stratification  is 
actually  centred  around  this  basic  notion,  whether  framed  as  a difference  between  organizational 
spaces  versus  occupational  spaces  in  labour  markets  (cf.  Maurice  et  al.  1986;  Muller  and  Shavit  1998; 


Brauns  et  al.  1998;  Jobert  et  al.  1997),  systems  of  internal  labour  markets  versus  systems  of 
occupational  labour  markets  (cf.  especially  Marsden  1986,  1990;  Marsden  and  Ryan  1995;  Eyraud  et 
al.  1990),  or  highly  versus  weakly  stratified  educational  systems  (Allmendinger  1989;  Hannan  et  al. 
1999). 

Now,  as  the  nature  of  the  stratification  system  determines  the  channelling  of  the  flows  of  individuals 
into  positions  (Kerckhoff  1995),  the  magnitude  of  such  institutional  effects  on  attainment  processes  is, 
for  two  main  reasons,  best  assessed  from  a cross-national  analysis  of  labour  market  entry:  first,  the 
basic  structure  of  education  and  training  systems  is  determined  at  a national  level,  so  that  only  cross- 
national analysis  will  provide  sufficient  institutional  variation  to  allow  these  type  of  effects  to  be 
detected.  Second,  labour  market  entry  provides  the  analyst  with  a “pure”  flow  situation  as,  per 
definition,  no  individual  is  already  allocated  to  a position.  Turning  the  argument  on  its  substantive  head 
again,  all  of  this  institutionalist  reasoning  implies  the  expectation  of  huge  cross-national  differences  in 
labour  market  entry  patterns,  closely  tied  to  the  nature  of  the  interlinkage  between  education  and 
training  systems  and  the  labour  market. 

Much  of  current  research  does  indeed  lend  considerable  support  to  the  above  arguments  (cf. 
Allmendinger  1989;  Kerckhoff  1995,  1996,  forthcoming;  Muller  and  Shavit  1998;  Shavit  and  Muller, 
forthcoming).  Still,  cross-national  comparisons  have  so  far  more  often  than  not  been  restricted  to 
analyses  of  a limited  number  (mostly  two)  out  of  a limited  set  of  (mostly  Northern  European  or 
Northern  American)  countries  (cf.  Muller  and  Shavit,  1998,  for  a notable  exception).  Drawing  on  a new 
database  for  the  countries  of  the  European  Union,  the  current  paper  is  able  to  provide  an  analysis 
addressing  the  variety  of  stratification  systems  across  a considerably  expanded  set  of  countries  and 
institutional  arrangements,  notably  including  Southern  Europe  among  the  country  cases.  Building  on 
data  for  twelve  member  states  of  the  European  Union,  the  paper  consequently  aims  to  put  the  above 
institutionalist  hypothesis  to  an  empirical  test:  do  we  empirically  observe  a dichotomy  of  stratification 
systems  shaping  labour  market  entry  in  European  countries?  And  furthermore,  do  country  contrasts 
conform  to  the  expectation  of  a dichotomy  in  terms  of  types  of  linkages  between  educational  systems 
and  labour  markets? 

In  order  to  pursue  these  questions,  the  paper  proceeds  to  develop  an  analytical  framework  for 
identifying  different  stratification  systems  from  their  empirical  consequences  rather  than  from 
institutional  analysis  (cf.  Marsden  1990;  Eyraud  et  al.  1990  for  related  earlier  attempts).  The  main 
institutional  argument  is  outlined  in  more  detail  in  the  following  section.  From  that,  a set  of  empirically 
testable  hypotheses  on  structural  differences  between  the  ideal  type  stratification  systems  in  terms  of 
patterns  of  labour  market  entry  is  delineated.  Section  3 then  proceeds  to  discuss  the  data  sources  and 
research  design  of  the  study,  as  well  as  the  operationalization  of  specific  hypotheses  gained  in 
Section  2.  After  that,  Section  4 presents  basic  descriptive  results  on  the  structure  of  labour  market 
entry  in  the  European  economies  under  study,  while  Section  5 discusses  the  main  empirical  analyses. 
The  findings  are  summarized  and  assessed  in  a concluding  section. 


ERiC 


4.3 


271 


2.  Institutional  Arrangements  and  Entry  Labour  Markets 

From  what  has  been  said  so  far,  the  core  institutionalist  claim  is  that  cross-national  similarities  and 
differences  in  the  transition  into  working  life  do  reflect  systemic  differences  in  the  sense  of  stemming 
from  the  operation  of  distinct  types  of  national  stratification  systems.  As  coined  by  Kerckhoff  (1995: 
342),  the  notion  of  stratification  systems  is  intended  to  address  distinct  types  of  channelling  the  flow  of 
individuals  to  positions.  Of  course,  a particular  stratification  system  is  always  the  consequence  of 
specific  institutional  arrangements,  i.e.  particular  institutional  interlocks  of  specific  types  of  education 
and  training  systems  with  specific  sets  of  labour  market  regulation  and  labour  market  policies,  or,  for 
that  matter,  a specific  style  family  formation  (Hannan  et  al.  1999;  cf.  the  general  argument  in  Soskice, 
forthcoming;  Hall  and  Soskice  1998).  With  respect  to  stratification  systems  in  industrial  societies,  the 
main  institutional  hypothesis  is  that  the  structure  of  the  education  and  training  system  is  a key  factor  in 
determining  the  nature  of  the  stratification  system,  resulting  in  two  distinct  institutional  equilibria  of 
particular  types  of  training  systems  and  thus  in  specific  patterns  of  stratification.  More  specifically,  it  is 
the  relative  reliance  of  market  matching  processes  on  formal  education  versus  experience  and 
mobility  which  is  at  stake  here.  In  the  context  of  vocationally  specific  and  occupationally  relevant  initial 
training,  the  education  and  training  system  performs  an  effective  presorting  of  individuals  and  allows 
for  a stratification  system  based  on  certified  skills.  In  the  absence  of  such  training  systems,  matching 
processes  have  to  rely  relatively  more  on  experience  and  mobility,  thus  yielding  a different  type  of 
stratification  system. 

What  then  should  be  the  mechanisms  bringing  about  such  differences  in  stratification?  As  the 
relationship  between  skills,  employment  relationships  and  labour  market  attainment  has  always  been 
a crucial  concern  of  labour  market  segmentation  theory  (cf.  Doeringer  and  Piore,  1971;  Sorensen  and 
Kalleberg,  1981),  it  may  come  as  no  surprise  that  the  most  lucid  description  of  the  operation  of  such 
stratification  systems  originates  from  that  research  tradition.  In  a segmentation  perspective,  the  main 
issue  is  the  ways  in  which  work  skills  are  provided  and  produced,  how  markets  for  skilled  labour  are 
organized  and  how  recruitment  into  more  skilled  positions  operates.  In  this  line  of  reasoning,  the  works 
by  David  Marsden  (1986,  1990,  1993,  1997;  cf.  also  Marsden  and  Ryan,  1995;  Eyraud  et  al.  1990) 
provide  probably  the  clearest  statement  of  an  institutional  theory  of  stratification  systems  in 
comparative  research.  As  most  other  research  can  easily  be  related  to  both  his  arguments  and 
terminology,  this  section  will  outline  the  argument  following  his  approach.  Marsden  phrases  the 
contrast  between  the  two  polar  stratification  systems  as  one  of  systems  of  internal  labour  markets 
(ILM  arrangements)  versus  systems  of  occupational  labour  markets  (OLM  arrangements).  The  key 
difference  between  the  two  is  the  presence  of  an  education  and  training  system  providing 
occupationally-specific  skills,  thus  transforming  the  stratification  system  from  an  ILM  type  into  an  OLM 
system  (Marsden,  1986,  holds  the  ILM  system  to  be  the  baseline  market  arrangement,  in  the  absence 
of  a sufficiently  specific  training  system). 

The  main  difference  in  the  operation  of  each  model  arrangement  then  is  in  the  institutional  rules  of 
access  to  skilled  worker  positions  (cf.  Figure  1 for  a graphical  sketch  of  the  reasoning),  which  are 


ERiC 


4.4 


27? 


conceptually  defined  as  positions  requiring  task-specific  skills  for  productive  work.  In  an  internal 
market  system,  available  educational  credentials  provide  little  guidance  in  allocating  individuals  and 
occupational  tasks  due  to  their  lack  of  occupational  specificity.  Therefore,  access  to  skilled  positions 
should  be  strongly  based  on  experience  and  mobility  criteria:  individuals’  potential  productivity  in 
particular  skilled  positions  can  only  be  assessed  from  employment  careers,  most  reliably  from  tenure. 
Similarly,  task-specific  skills  are  regularly  provided  as  firm-specific  skills,  again  strongly  linking  the 
allocation  of  skilled  positions  to  work  experience.  In  that  sense,  skilled  positions  are  relatively 
sheltered  both  from  external  market  competition  and  from  competition  with  labour  market  entrants. 
Skilled  positions  thus  exist  as  a sheltered  internal  segment  of  the  labour  market,  while  an  external 
segment  provides  lower  skilled  employment  and  entry  ports  into  firm  internal  labour  markets. 

In  contrast  to  this,  occupational  market  systems  provide  a quite  different  mechanism  for  attaining 
skilled  worker  positions:  in  this  context,  skilled  positions  are  available  to  the  external  market,  yet 
competition  for  these  is  restricted  to  those  workers  with  the  appropriate  occupational  skills.  In  this 
model,  applicants  have  been  presorted  into  specific  occupational  sub-markets  where  recruitment  for 


Figure  1 Models  of  Internal  and  Occupational  Labour  Markets 


Internal  Labour  Market 


Occupational  Labour  Market 


Skilled 

labour 


Semi-skilled 

labour 


Unskilled 

labour 


Main  inflows 


•Main  outflows 


Source:  Marsden  (1993) 


O 

ERIC 


4.5 


273 


skilled  positions  takes  place.  Given  such  a training  system  delivering  occupationally  specific  certified 
skills,  available  credentials  in  this  case  provide  a sound  basis  for  allocating  workers  and  positions 
productively.  Therefore,  there  is  comparatively  little  need  to  resort  to  allocation  mechanisms  relying  on 
experience  or  mobility.  After  all,  the  OLM  model  is  thus  argued  to  operate  with  a strong  orientation 
towards  external  labour  markets  and  inter-firm  mobility  in  allocating  skilled-level  positions  (cf.  Marsden 
1986,  1990,  1993;  Eyraud  etal.  1990). 


The  Consequences  of  Stratification  Systems  for  Labour  Market  Entry 

Education,  work  experience  and  mobility  provide  the  means  of  labour  allocation  - with  the  foregoing 
discussion  positing  the  existence  of  two  distinct  institutional  equilibria  relying  either  primarily  on  skill- 
based  or  strongly  on  experience-based  mechanisms.  As  such,  this  institutional  hypothesis  does  have 
a serious  implication  for  labour  market  entry  in  different  stratification  systems:  on  the  one  hand,  one 
expects  to  observe  strong  effects  of  differences  in  stratification  systems  as  initially  all  individuals  are 
newly  allocated  to  positions.  On  the  other  hand,  and  more  importantly,  if  the  matching  between 
persons  and  jobs  to  varying  degrees  relies  on  skills  or  experience,  then  the  allocation  outcomes  of 
those  lacking  experience  should  differ  markedly  across  systems.  In  the  terminology  of  transition 
research,  this  is  expressed  in  descriptions  such  as  OLM  arrangements  allowing  for  a structured  labour 
force  integration  in  the  sense  of  a strict  channelling  of  individuals  into  positions  by  education  and  an 
immediate  close  match  between  qualifications  and  labour  market  positions.  In  contrast,  labour  market 
entry  in  ILM  systems  is  much  less  tightly  structured  by  education,  less  orderly,  more  amenable  to 
career  contingencies  and  firm  behaviour.  In  a sense,  the  available  supply  of  qualifications  necessitates 
a gradual  integration  into  the  labour  force  by  achieving  job-skill  matches  primarily  via  work  experience 
and  mobility  rather  than  initial  skills  (cf.  Kerckhoff  1995;  Muller  and  Shavit  1998;  Marsden  and  Ryan 
1995;  but  e.g.  also  the  terms  regulated  integration  versus  selective  exclusion  by  Garonna  and  Ryan, 
1989). 

This  institutional  contrast  can  actually  be  restated  in  terms  of  two  basic  premises  that  allow  for 
generating  a set  of  more  detailed  hypotheses  on  consequences  of  stratification  systems  - which  can 
then  be  operationalized  for  use  in  the  empirical  analyses  later  on.  These  premises  relate  to  a 
structural  difference  in  terms  of  both  labour  market  exclusion  and  positional  attainment,  and  may  be 
stated  as  follows: 

(A)  Exclusion  Effect: 

Labour  markets  vary  in  the  evaluation  of  “least  desirability”.  An  ILM  arrangement  implies  a 
relatively  more  positive  evaluation  of  experience,  sheltering  the  adult  work  force  from 
competition  with  labour  market  entrants.  Therefore,  labour  market  exclusion  exhibits  a 
strong  negative  bias  towards  market  entrants.  In  contrast,  market  exclusion  in  OLM 
contexts  can  be  expected  to  operate  mostly  through  lack  of  skills  rather  than  experience. 


ERIC 


4.6  274 


(B) 


Attainment  Effect: 


Average  initial  job-person  match  quality  is  lower,  but  increases  with  labour  force 
experience  in  ILM  systems  while  OLM  arrangements  provide  the  base  for  an  experience- 
constant  match  quality.  As  such,  the  scope  for  any  type  of  labour  market  mobility  - as  a 
major  means  of  match  adjustment  (cf.  Jovanovic  1979;  LeGrand  and  TShlin  1998)  - is 
expected  to  be  larger  in  ILM  systems. 


From  these  basic  tenets,  a set  of  more  specific  implications  for  labour  market  entry  in  different 
stratification  systems  follows,  as  will  be  outlined  briefly  (Table  1 provides  an  overview  of  these). 
Turning  first  to  the  structure  of  externally  recruited  employment  positions,  a crucial  element  of  the  ILM 
arrangement  is  that  such  outsiders,  including  labour  market  entrants,  are  recruited  mainly  into  lower 
skilled  positions,  at  least  in  comparison  to  the  broader  dispersion  of  contracting  across  skill  levels 
enabled  by  OLM  systems.  Given  the  expected  experience  bias  in  ILM  exclusion  patterns  (cf. 
assumption  A),  one  can  furthermore  predict  a contrast  between  either  arrangement  in  terms  of  the 
experience  grading  of  exclusion  risks.  Taking  a broader  view  on  exclusion,  one  can  therefore  expect 
that  both  unemployment  and  lower-skilled  employment  is  much  more  concentrated  among  market 
entrants  in  ILM  systems  than  in  OLM  contexts.  In  turn,  drawing  on  the  expected  contrast  in  terms  of 
qualifications,  one  can  expect  a stronger  qualificational  stratification  in  initial  exclusion  risks  in  OLM 
systems  than  in  an  ILM  arrangement. 

Based  on  the  assumptions  about  early  career  adjustment  processes  (assumption  B),  one  can 
moreover  arrive  at  a set  of  additional  institutional  predictions  related  to  job  mobility  and  career  patterns 
of  attainment.  From  the  notion  that  early  career  adjustment  processes  are  necessitated  by  an  ILM 
system  while  OLM  systems  allow  for  effective  matches  already  at  market  entry,  one  can  predict  the 
following  contrasts:  first,  there  will  be  an  experience-graded  pattern  of  job  mobility  in  ILM  systems 
which  will  be  absent  in  OLM  contexts.  As  far  as  both  arrangements  ultimately  yield  similarly  effective 
assignments,  later  career  mobility  rates  will  be  similar  and,  in  consequence,  initial  mobility  rates 
comparatively  higher  in  ILM  systems.  Apart  from  mobility  patterns,  career  attainment  profiles  can  be 
expected  to  look  quite  different  for  the  two  model  arrangements.  As  ILM  systems  have  to  rely  on  firm- 
internal  provision  of  training  and  internal  promotion  to  more  skilled  positions,  there  is,  secondly,  a 
much  greater  role  for  occupational  and  positional  upgrading  over  early  labour  market  careers  in  these 
systems.  Thus,  a substantial  experience  effect  on  attainment  is  expected  for  ILM  arrangements  which 
should  be  absent  in  an  OLM  context.  As  occupational  upgrading  may  be  expected  to  be  stratified  by 
formal  qualifications,  this  mechanism  can  be  expected  to  lead  to  increasing  differences  in 
qualificational  attainment  with  experience  in  ILM  systems.  Finally,  as  this  mechanism  by  definition 
depends  on  firm-specific  arrangements,  this  can  be  expected  to  imply  increasing  individual  variation  in 
attainment  patterns  with  experience  in  ILM  systems,  which  is  part  of  the  more  flexible  nature  of  this 
integration  regime. 


O 

ERLC 


4.7 


275 


Table  1 


Models  Contrasted:  Structural  Differences  between  Market  Arrangements 


ILM  Arrangement 

OLM  Arrangement 

Labour  Mobility 

Extent  of 

Labour  Mobility  in  Early  Career 

Relatively  high 

Relatively  low 

Experience  Structure 
of  Labour  Mobility 

concentration  on  entrants 

dispersion 

across  experience  groups 

Labour  Market  Positions 

Positional  Structure 
of  External  Recruitment 

Concentrated 
on  lower  skilled  positions 

Dispersion  across 
different  levels  of  skills 

Experience  Structure 
of  Lower  Skilled  Employment 

Exclusive, 

risk  concentration  on  entrants 

Inclusive,  risk  dispersion 
across  experience  groups 

Qualificational  Structure 
of  Lower  Skilled  Employment 
at  Market  Entry 

Inclusive,  risk  dispersion  across 
qualificational  backgrounds 

Exclusive, 

risk  concentration  on  least  qualified 

Experience  Structure 
of  Unemployment 

Exclusive, 

risk  concentration  on  entrants 

Inclusive,  risk  dispersion 
across  experience  groups 

Qualificational  Structure 
of  Unemployment  at  Market  Entry 

Inclusive,  risk  dispersion  across 
qualificational  backgrounds 

Exclusive, 

risk  concentration  on  least  qualified 

Career  Patterns  of  Attainment 

Payoff  Progression 

with  Labour  Force  Experience 

Substantial 

Flat 

Initial  Educational  Payoff  Premium 

Low 

Strong 

Evolution  of  Payoff  Differential  with 
Labour  Force  Experience 

Increasing  differentials  with 
labour  market  experience/tenure 

Constant  differentials 
with  labour  market  experience 

Payoff  Dispersion 
within  Skill  Groups 

Increasing  dispersion  with 
labour  market  experience/tenure 

Constant  dispersion 
with  labour  market  experience 

3.  Methodological  Approach  and  Institutional  Hypotheses 

The  following  analyses  aim  to  address  the  analytical  power  of  the  institutional  contrast  between  ILM 
and  OLM  stratification  systems  in  a cross-national  analysis  of  labour  market  entry  patterns  for  twelve 
countries  of  the  European  Union.  The  analyses  draw  on  data  from  the  European  Community  Labour 
Force  Survey,  a standardised  database  compiled  from  the  national  Labour  Force  Surveys  in  the 
member  countries  of  the  European  Union  (cf.  Eurostat  1996  for  details).1  The  main  advantage  of  the 
database  is  its  coverage  of  a broad  set  of  European  countries  with  divergent  institutional 
arrangements  in  educational  systems  and  labour  markets.  Furthermore,  Labour  Force  Surveys  are 


O 

ERIC 


The  data  has  kindly  been  provided  by  EUROSTAT,  the  Statistical  Office  of  the  European  Union.  Of  course, 
EUROSTAT  is  not  responsible  for  the  use  of  the  data,  the  interpretations  drawn,  nor  the  views  held  by  the 
author. 


4.8 


?7  ft 


geared  towards  comparative  stratification  analysis  as  they  are  conducted  within  regular  survey  periods 
applying  standardised  instruments  to  large  survey  sample  sizes.  In  this  paper,  data  for  the  period 
between  1992  and  1997  will  be  used.2 

For  the  analyses,  a sub-sample  of  individuals  in  early  career  stages  is  drawn,  namely  all  individuals  in 
the  labour  force  with  no  more  than  10  years  of  labour  force  experience,  having  attained  no  more  than 
ISCED  level  3 qualifications3  and  not  participating  in.  formal  education  and  training.4  In  passing,  it  is 
noted  that  employment  status  is  measured  according  to  the  international  standard  definitions  (cf.  ILO 
1990),  with  the  exception  that  participation  in  formal  education  and  training  is  given  priority  status  to 
the  ILO  classification.  As  such,  working  pupils  or  students,  but  also  apprentices  or  individuals  in  similar 
training  environments  are  not  considered  as  part  of  the  active  labour  force.  Secondly,  labour  force 
experience  is  not  measured  in  the  data,  but  rather  proxied  as  potential  experience,  i.e.  years  since  last 
leaving  education  and  training.  The  analyses  apply  typical  graduation  ages  as  provided  by  OECD 
(1997).  Labour  market  attainment  is  finally  measured  by  three  different  concepts.  The  first  one, 
unemployment,  naturally  follows  from  the  ILO  definitions  applied.  Furthermore,  information  on 
individual  occupation  is  used  to  assess  the  nature  of  job  positions.  Occupations  enter  the  analyses  at 
two  points:  this  information  is  used  to  define  lower-skilled  employment  positions  (cf.  details  in  Table  2) 
and  to  construct  an  index  of  occupational  status,  applying  the  ISEI  occupational  status  scale  as 
developed  by  Harry  Ganzeboom  and  colleagues  (cf.  Ganzeboom  et  al.  1992;  Ganzeboom  and 
Treiman  1996;  Wegener,  1992,  provides  a general  discussion  of  status  scales). 

To  assess  cross-national  differences  in  the  stratification  of  initial  labour  market  outcomes,  the  relation 
of  labour  market  attainment  to  education  and  experience  is  examined  according  to  the  expected 
structural  contrasts  between  the  institutional  model  arrangements  as  detailed  in  Table  1 above.  To 
describe  each  aspect  of  labour  market  attainment  considered,  a set  of  auxiliary  regressions  of  the 
format 

(1)  Y = b0  + b^In  experience)  + b2(intermed.  skills)  + b3(ln  experience  * intermed.  skills) 

is  estimated  for  each  country  and  year  in  order  to  provide  a measurement  of  skill  and  experience 
effects.5  Added  to  this  set  of  regression  parameters,  two  simple  proportion  measures  describe  the 
overall  extent  of  labour  mobility  and  the  relative  importance  of  lower-skilled  employment  in  external 


2 Because  of  small  sample  sizes  and  resulting  unstable  detailed  estimates,  Luxembourg  is  excluded  from  this 
study.  Also,  Sweden  and  Finland  will  be  excluded  from  the  analyses  of  the  paper  as  complete  data  were 
available  for  1997  only. 

3 ISCED  level  3 corresponds  to  the  completion  of  upper  secondary  education  or  complementary  vocational 
training.  Thus,  individuals  having  obtained  post-secondary  or  tertiary  qualifications  are  excluded.  This 
restriction  is  due  to  the  expectation  already  implicit  in  the  above  discussion  of  market  systems  that  cross- 
national variation  is  expected  to  apply  primarily  at  the  level  of  intermediate  skills  (Marsden  1990;  Marsden  and 
Ryan  1995;  Muller  and  Shavit  1998;  Hannan  et  al.  1999). 

4 Participation  in  formal  education  and  training  is  specified  as  participation  in  initial  training  or  training  for  other 
purposes  if  enrolment  is  either  in  general  secondary  level  education  tracks,  tracks  at  vocational  schools  of  at 
least  one  year  duration,  dual  system  training  or  tertiary  level  studies. 


4.9 


277 


recruitment.  Table  2 below  provides  full  details  of  the  estimates  gained,  linking  these  to  the  expected 
structural  contrasts  between  ILM  and  OLM  arrangements.  In  sum,  the  set  of  auxiliary  regression 
results  in  an  estimate  of  11  indicators  in  total,  measuring  different  aspects  of  the  stratification  of  initial 
labour  market  experiences  in  our  set  of  twelve  European  countries. 


Table  2 Models  Contrasted:  Empirical  Indicators  of  Labour  Market  Structure 


Labour  Mobility 

Labour  Mobility  Mobility  Rate:  Proportion  of  employer  or  employment  status  change  over  the 

in  Early  Career  last  year  among  last  year's  labour  force 

Experience-Grading:  Experience  effect  bi  on  mobility  rate 


Labour  Market  Positions 


Structure  Lower  Skill  Bias  of  Recruitment: 

of  Recruitment  Behaviour  Ratio  of  the  proportion  of  lower  skilled  employment  among  external  recruitments  to 

the  proportion  of  unskilled  employment  among  all  other  employment1 


Structure  of 
Secondary  Labour 
Market 


Experience-Grading:  Experience  effect  bi  on  lower  skilled  employment  rate 

Qualificational  Grading:  Qualification  effect  b2  on  lower  skilled  employment  rate 


Structure  of 
Unemployment  in  Early 
Career 


Experience-Grading:  Experience  effect  bi  on  unemployment  rate 

Qualificational  Grading:  Qualification  effect  b2  on  unemployment  rate 


Career  Patterns  of  Attainment 


Payoff  Progression 


Payoff  Differential 
between  Skill  Groups 

Payoff  Differential 
Progression 

Payoff  Dispersion 
within  Skill  Groups 


Experience-Grading  of  Attainment: 

Experience  effect  bi  on  occupational  status  attainment 

Qualificational  Grading  of  Attainment: 

Qualification  effect  b2  on  occupational  status  attainment 

Interaction  of  Experience  and  Qualifications  on  Attainment: 

Interaction  effect  b3  of  experience  and  qualifications  on  occupational  status  attainment 

Experience-Grading  of  Attainment  Dispersion: 

Experience  effect  bi  on  dispersion  of  occupational  status  attainment 


Notes: 

Parameters  are  gained  from  the  following  regression  of  respective  macrolevel  relations: 

Y = bo  + bi(ln  experience)  + b2(intermed.  qualifications)  + b3(interaction  In  experience-qualifications) 
Dependent  variables  were  labour  force  mobility  rates,  rates  of  lower  skilled  employment,  unemployment  rates  and 
attainment  levels  and  dispersion  in  terms  of  ISEI  scores  (cf.  Figures  2-6  for  partial  graphical  displays  of  the  data 
used). 

Lower  qualifications  - ISCED  levels  0-2  (max.  lower  secondary  education) 

Intermediate  qualifications  - ISCED  level  3 (upper  secondary  education) 

Lower-skilled  employment  includes  categories  421,  422,  512,  516-522,  611-615,  822-830,  832-933  of  ISC088- 
COM  (cf.  Eurostat  1996) 


In  one  case,  the  setup  of  the  auxiliary  models  differs:  the  mobility  equation  only  controls  for  the  main 
experience  effect.  All  other  equations  include  ail  three  effects,  even  if  not  all  parameters  are  considered  later 
on. 


4.10 


27ft 


Based  on  this  setup,  the  analyses  then  attempt  to  empirically  identify  the  theorized  distinct 
stratification  systems  from  and  for  the  set  of  the  countries  under  study.  Keeping  the  notion  of  systemic 
features  in  mind,  it  is  obvious  that  the  expectation  is  to  find  two  sets  of  clearly  distinct  stratification 
systems,  tending  towards  the  features  of  an  ILM  or  OLM  system,  respectively.  To  identify  such  distinct 
systems,  the  chosen  indicators  must  be  assessed  simultaneously  in  order  to  form  groups  of  country- 
year  cases  which  exhibit  consistent  similarities  and  dissimilarities  on  the  set  of  structural  contrasts, 
rather  than  singular  deviations  on  some  dimensions.  To  achieve  this  goal,  cluster  analyses  are 
performed  on  the  country-year  cases  in  the  sample.  Effectively,  both  the  “clusterability"  of  countries 
themselves  as  well  as  the  substantive  differences  between  groups  of  countries  provide  an  empirical 
assessment  of  the  institutional  claims  advanced  so  far.  While  the  notion  of  clusterability  obviously 
relates  to  the  issue  of  fit  between  the  cluster  solution  and  real  data,  the  consistency  of  the  substantive 
differences  in  terms  of  stratification  outcomes  between  country  clusters  and  theoretical  expectations  is 
finally  assessed  by  means  of  discriminant  analysis. 

As  indicated  earlier,  this  study  builds  on  a rich  set  of  previous  analyses  (cf.  Hannan  et  al.  1997,  1999; 
Muller  and  Shavit  1998;  Shavit  and  Muller,  forthcoming;  Allmendinger  1989  among  others).  Naturally,* 
the  formation  of  hypotheses  concerning  the  classification  of  European  countries  in  terms  of  the  above 
framework  draws  heavily  on  this  research.  Ultimately,  the  expectation  is  that  European  countries 
adhere  to  one  pole  of  either  ILM  or  OLM  stratification  systems.  As  the  structure  of  the  education  and 
training  system  has  to  be  regarded  as  a major  prerequisite  for  the  development  of  each  market 
arrangement,  a classification  hypothesis  can  essentially  be  based  on  the  vocational  specificity  of  the 
training  system  at  the  upper  secondary  level.  Most  observers  have  arrived  at  a basic  distinction 
between  educational  systems  focusing  on  the  provision  of  school-based  general  education  contrasted 
to  those  emphasising  vocationally  specific  training,  typically  provided  in  the  form  of  apprenticeship  (cf. 
Hannan  et  al.  1997,  1999;  Muller  and  Shavit  1998;  Shavit  and  Muller,  forthcoming).  In  general,  the 
latter  systems  are  to  be  associated  with  OLM  arrangements  as  they  provide  specific  skills  already  for 
those  entering  the  labour  market,  while  the  former  systems  should  tend  towards  ILM  systems.  Given 
the  set  of  countries  under  study,  this  leads  to  expect  a classification  of  Austria,  Denmark,  Germany, 
and  the  Netherlands  as  OLM  systems  as  these  countries  have  extensive  apprenticeship  systems 
and/or  extensively  occupationally  differentiated  school-based  training  systems  (cf.  the  overview  in 
OECD  1997,  Hannan  et  al.  1999).  According  to  the  structure  of  education  and  training  systems,  the 
other  EU  countries  should  form  a cluster  of  ILM  countries. 


Bearing  in  mind  results  from  the  above  and  related  earlier  studies,  there  are  some  particularly 
interesting  country  cases  included  in  the  analysis:  for  example,  the  estimated  position  of  Denmark  and 
the  Netherlands  will  be  of  particular  interest  as  vocational  training  is  much  more  school-based  as 
compared  to  the  traditional  dual-system  countries  of  Austria  and  Germany.  On  the  other  hand,  the 
position  of  the  UK  along  the  ILM-OLM  axis  has  been  an  issue  of  some  debate  as  some  researchers 
claim  a near-OLM  context  (e.g.  Marsden  1990;  Kerckhoff  1995),  while  others  argue  strongly  against 
such  an  idea  (cf.  Soskice  1993).  Furthermore,  except  for  Marsden's  (1990)  result  of  Italy  belonging  to 
the  ILM  model,  there  is  little  systematic  evidence  on  Southern  Europe  at  all.  In  a sense,  the  empirical 


4.11 


279 


results  are  thus  likely  to  indicate  critical  threshold(s)  of  the  vocational  specificity  of  training  systems  for 
transforming  the  stratification  system  into  an  OLM  model.  For  now,  the  following  section  starts  the 
presentation  of  empirical  findings  on  labour  market  entry  in  EU  countries  in  a descriptive  fashion. 


4.  The  Structure  of  Labour  Market  Entry  across  Europe 

The  following  aims  to  provide  a descriptive  overview  of  some  core  structural  features  of  labour  market 
entry  in  the  set  of  European  countries.  Since  much  of  the  theoretical  argument  amounts  to  expecting 
major  cross-national  differences  in  the  role  of  labour  force  experience  in  the  attainment  process  at  the 
level  of  intermediate  skills,  the  following  descriptive  presentation  will  mostly  focus  on  results  for  this 
group.6  In  terms  of  substantive  issues,  the  section  addresses  cross-national  differences  in  the  rate  of 
labour  market  mobility,  the  incidence  of  unemployment  and  lower  skilled  employment  as  well  as  early 
career  status  attainment  patterns  of  labour  market  entrants  with  intermediate  skills  across  European 
Union  countries.  The  final  part  of  this  section  will  then  present  a more  thorough  discussion  of  cross- 
national similiarities  and  differences  in  terms  of  the  set  of  structural  indicators  developed  above.  It  is 
noted  in  passing  that  all  descriptive  evidence  presented  in  this  section  refers  to  average  estimates  for 
each  country  during  the  period  1992-1997,  while  country-year  cases  will  be  used  in  the  analyses  of 
Section  5 below. 


4.1  Unemployment  and  Volatility  in  Early  Careers 

As  a first  indicator  for  cross-national  differences  in  stratification  systems,  Figures  2 and  3 below 
provide  evidence  on  unemployment  risks  and  the  volatility  of  initial  labour  market  positions  across  the 
twelve  EU  countries  under  study.  For  the  purpose  of  presentation,  three  sets  of  countries  are 
tentatively  grouped  together  in  these  and  the  following  figures,  consisting  of  the  expected  set  of  OLM- 
type  countries  (Austria,  Denmark,  Germany,  the  Netherlands),  the  Northern  European  ILM-type 
countries  (Belgium,  France,  Ireland,  the  United  Kingdom),  and,  finally,  Southern  Europe  (Greece,  Italy, 
Spain,  Portugal). 

Focusing  on  unemployment  patterns  first,  Figure  2 below  shows  clear  evidence  of  strong  cross- 
national differences  among  EU  countries.  At  the  level  of  intermediate  qualifications  - equated  with 
ISCED  3 level  education  here  - there  is  an  obvious  contrast  in  unemployment  experiences  between 
young  people  in  the  set  of  potential  OLM  system  countries  in  the  left  panel  as  compared  to  all  other 
EU  countries.  In  Austria,  Denmark,  Germany  and  the  Netherlands,  this  group  of  labour  market 
entrants  face  considerably  lower  unemployment  risk,  than  their  counterparts  in  other  Western, 
Southern  or  Northern  European  countries. 


Extended  descriptive  results  are  available  from  the  author  on  request. 


O 

ERIC 


4.12 


280 


Figure  2 Unemployment  Rates  and  Labour  Force  Experience 


CM 

OO 

CM 


CO  © 

c 

© 

f-  b. 


CO  UJ 


CO  © 

c 

© 

Is*  'C 
© 
a 

X 

CO  UJ 

© 

o 

CO  * 


c 


CO 
X3 
v-  C 
-C  TO 
ca  n 

II 

CD  CD 

o z 

{{ 


.55  co 


3 CD 

< O 

M 


CO 

CO 

c 

o 

TO 

O) 

2 

'^3 

TO 

o 

S* 

c 

TO 

b 

o 

'c 

E 

.c 

O 

o 

to 

nT 

D) 

o> 

O 

o> 

>* 

1 

-Q 

CM 

CO 

(D 

O) 

o> 

r~ 

TO 

E 

r 

TO 

C 

8 

*g> 

o 

*c 

o 

u. 

. , 

o 

o> 

o 

jQ 

c 

TO 

£ 

-TO 

o 

o 

E 

C 

TO 

CO 

c 

E 

s 

(D 

o 

CO 

TO 

O 

u_ 

Q. 

TO 

w 

c 

TO 

TO 

CO 

§• 

TO 

2 

c 

TO 

—1 

UJ 

co 

Q> 

2 

3 

O 

CO 


OO 

C\f 


o 

ERIC 


Figure  3 Job  Mobility  and  Labour  Force  Experience 


Notes:  Lines  represent  smoothing  of  original  estimates  by  logarithmic  functions 

Sources : European  Community  Labour  Force  Survey  1992-1997 \ country  averages 


The  most  interesting  contrast  is  less  in  the  differences  in  overall  levels  but  rather  the  extent  to  which 
these  differences  converge  over  the  first  years  in  the  labour  market.  After  a period  of  ten  years, 
countries  appear  substantially  less  heterogeneous  in  terms  of  unemployment  risks:  this  is  due  to  the 
fact  that  in  all  countries  except  those  in  the  left  panel,  a substantial  experience-grading  of 
unemployment  risks  is  observed.  That  is,  in  the  majority  of  European  countries  unemployment  is 
concentrated  among  labour  market  entrants,  yet  unemployment  risks  wear  off  with  time  in  the  labour 
force.  Thus,  cross-national  variation  is  both  arguably  strongest  immediately  at  labour  market  entry  and 
largely  resolved  with  increasing  labour  force  experience. 

Does  this  result  also  imply  that  we  observe  little  mobility  and  volatility  in  early  careers  in  the  set  of 
potential  OLM  systems?  Not  exactly  so,  according  to  the  results  on  labour  market  mobility  rates  given 
in  Figure  3.  Instead  of  a clear-cut  contrast  between  the  left  panel  of  Figure  3 as  compared  to  both  the 
middle  and  right  panels,  a pattern  common  to  all  Northern  European  countries  emerges.  In  these 
countries,  initial  employment  is  apparently  much  less  stable  than  employment  at  later  career  stages.  In 
the  initial  career  stage,  up  to  two  thirds  of  young  people  in  employment  change  employment  and/or 
employment  status  within  one  year.  This  proportion  is  reduced  to  approximately  20%-30%  over  the 
first  10  years  in  the  labour  market.  More  detailed  results  in  fact  show  some  variation  among  Northern 
European  countries  in  the  relative  importance  of  employment  versus  employment  status  changes  in 
generating  this  overall  mobility  rate,  which  is  consistent  with  the  view  that  job-to-job  rather  than  job-to- 
unemployment-to  job  transitions  dominate  the  picture  in  the  OLM-type  countries  of  the  left  panel.  Still, 
the  similarity  in  terms  of  overall  volatility  levels  and  patterns  of  experience-grading  is  impressive  and 
unexpected  from  theoretical  considerations  about  ILM/OLM  contrasts.  Indeed,  the  major  deviating 
cases  are  the  Southern  European  countries,  with  the  exception  of  Spain,  where  volatility  of  labour 
market  positions  is  low  even  at  the  outset  of  labour  market  careers. 


4.2  Lower-Skilled  Employment  and  Career  Patterns  of  Attainment 

Apart  from  volatility  in  early  career  phases  in  terms  of  job  mobility  and  unemployment  risks,  it  is  the 
nature  of  initial  employment  and  the  direction  of  occupational  mobility  that  is  important  to  the 
institutional  argument  assessed  here.  Specifically,  it  is  of  interest  to  see  whether  the  structure  of  entry 
ports  differs  across  countries. 

As  first  evidence  of  this,  Figure  4 shows  results  on  the  incidence  of  lower-skilled  employment  in 
relation  to  labour  force  experience.  According  to  this  indicator,  there  is  one  common  pattern  of  labour 
market  entry  for  the  broad  majority  of  European  countries:  in  almost  all  countries,  between  50%-60% 
of  market  entrants  hold  such  lower-skilled  positions  at  the  very  outset  of  careers,  with  this  percentage 
being  reduced  afterwards  by  around  20%  in  ten  years.  This  decline  is  most  pronounced  in  early  years 
in  the  labour  force,  indicating  that  major  occupational  upgrading  occurs  over  the  initial  employment 
career.  Among  the  twelve  EU  countries,  however,  Austria  and  Germany  and  in  part  Italy  stand  apart, 
exhibiting  absent  to  weak  effects  of  experience  on  the  incidence  of  lower-skilled  employment.  Again, 


0 


Figure  4 Lower-Skilled  Employment  and  Labour  Force  Experience 


F- 

QO 

C\t 


CD 

oo 


Notes:  Lines  represent  smoothing  of  original  estimates  by  polynomial  or  logarithmic  functions 

Sources:  European  Community  Labour  Force  Survey  1992-1997,  country  averages 


Figure  5 Status  Attainment  (ISEI  score)  and  Labour  Force  Experience 


03 

00 

CM 


V) 

c 

o 

o 

c 

3 

v»— 

o 

E 

CO 

x: 

05 

O) 

'u 

TO 

5 

CD 

05 

O 

2* 

CO 

o 

£ 

To 

c 

E 

a 

o 

o 

o 

c 

nT 

> 

o 

05 

Q. 

O) 

>. 

1 

-Q 

CNJ 

(/) 

0) 

O) 

05 

Y"» 

TO 

E 

2 

to 

0) 

3 

CO 

15 

c 

05 

O 

O) 

o 

o 

0. 

M— 

w. 

o 

O) 

o 

JQ 

c 

CO 

£ 

^1 

o 

o 

E 

c 

o 

(/) 

E 

c 

£ 

0) 

o 

w 

05 

O 

u. 

Q_ 

05 

t_ 

c 

03 

05 

V) 

05 

1 

C 

3 

— 1 

UJ 

V) 

. . <u 

10  p 

£ 3 

o o 

Z CO 


00 

oo 

CM 


o 

ERLC 


the  overall  proportion  of  lower-skilled  employment  after  some  five  to  ten  years  is  quite  similar  to  that 
found  for  other  European  countries.  The  cross-national  difference  is  again  one  of  early  careers, 
although  it  is  worth  stressing  the  difference  between  Austria  and  Germany  on  the  one  hand  and 
Denmark  and  the  Netherlands  on  the  other  in  this  respect. 

Such  occupational  upward  mobility  should  naturally  also  be  reflected  in  patterns  of  status  attainment. 
Figure  5 presents  the  outcomes  of  this  analysis,  depicting  the  relation  of  ISEI  status  attainment  to 
labour  force  experience.  Again  only  marginal  cross-national  differences  appear:  the  broad  majority  of 
European  countries  - all  countries  in  the  middle  and  right  panels,  including  Denmark  and  maybe  the 
Netherlands  from  the  left  panel  - show  a pattern  of  gradually  increasing  occupational  status.  Across 
the  board,  average  status  gains  for  ISCED  3 leavers  are  between  4-6  points  over  the  initial  ten  years 
in  the  labour  market.  The  only  exceptions  to  this  rule  are  again  Austria  and  Germany,  where 
occupational  attainment  patterns  exhibit  flatter  slopes  of  approximately  2 ISEI  points  in  ten  years. 
Apparently,  there  is  again  some  variation  in  this  respect  within  the  countries  of  the  left  panel,  with 
Denmark  and  the  Netherlands  slightly  deviating  from  the  Austrian-German  pattern.  Such  occupational 
upward  mobility  should  naturally  also  be  reflected  in  patterns  of  status  attainment.  Figure  5 presents 
the  outcomes  of  this  analysis,  depicting  the  relation  of  ISEI  status  attainment  to  labour  force 
experience.  Again  only  marginal  cross-national  differences  appear:  the  broad  majority  of  European 
countries  - all  countries  in  the  middle  and  right  panels,  including  Denmark  and  maybe  the  Netherlands 
from  the  left  panel  - show  a pattern  of  gradually  increasing  occupational  status.  Across  the  board, 
average  status  gains  for  ISCED  3 leavers  are  between  4-6  points  over  the  initial  ten  years  in  the 
labour  market.  The  only  exceptions  to  this  rule  are  again  Austria  and  Germany,  where  occupational 
attainment  patterns  exhibit  flatter  slopes  of  approximately  2 ISEI  points  in  ten  years.  Apparently,  there 
is  again  some  variation  in  this  respect  within  the  countries  of  the  left  panel,  with  Denmark  and  the 
Netherlands  slightly  deviating  from  the  Austrian-German  pattern. 


4.3  Summary:  Country  Differences  in  the  Set  of  Structural  Indicators 

Having  thus  briefly  discussed  some  core  descriptive  results  and  provided  some  flavour  of  the  data 
used,  I now  turn  to  a description  of  cross-national  differences  in  labour  market  entry  patterns  in  terms 
of  the  structural  indicators  developed  to  identify  the  operation  of  the  two  distinct  hypothesised 
stratification  systems.  As  a summary  of  the  country  scores  and  country  differences  in  these  indicators, 
Figure  6 below  provides  the  country  scores  on  the  set  of  indicators  averaged  over  the  available  years. 
The  data  table  in  the  lower  part  of  Figure  6 reports  the  original  scores,  while  the  graph  in  the  upper 
part  of  Figure  6 represents  the  z-standardised  scores  which  will  mostly  be  relied  on  for  country 
comparison. 

Taking  a look  at  specific  indicators,  the  two  parameters  describing  labour  market  mobility  rates  in  early 
careers  in  the  twelve  countries  reiterate  the  results  already  reported  above,  slightly  rephrased  in  terms 
of  the  chosen  operationalisation.  Judged  from  the  z-standardised  scores  depicted  in  the  graph,  the 


Figure  6 Structural  Indicators  for  European  Entry  Labour  Markets 


CM 

o> 

CM 


to 

CM 

CM 


to 

r*» 


to 

N- 


tO 

CM 

o 


to 

CM 


to 


to 

CM 


to 


to 

CM 


DC 

0 

* 

(/) 

LU 

t 

* 

ID 

1 

LU 

9= 

? 

DC 

LL 

{ 

LU 

CQ 

} 

—I 

Z 

\ 

* 

o 

LU 

Q 

+ 

I- 

< 

\ 


ERiC 


Notes:  Graph  represents  z-standardised  indicator  scores;  original  scores  are  provided  in  the  table. 

Sources:  European  Community  Labour  Force  Survey  1992-1997,  country  averages 


contrast  between  substantial  mobility  in  all  Northern  European  countries  versus  low  mobility  in  all 
Southern  European  countries  except  Spain  is  again  immediately  evident.  There  is  little  in  the  data  to 
suggest  that  mobility  rates  are  lower  for  the  set  of  hypothesised  OLM  systems,  not  even  as  compared 
to  their  Northern  European  counterparts.  Rather,  the  expected  OLM  systems  of  Denmark  and  the 
Netherlands  exhibit  the  highest  mobility  rates  of  all  countries  under  study.  In  addition,  the  obvious 
result  of  strong  experience  effects  on  these  rates  in  all  Northern  European  countries  versus  largely 
absent  experience  effects  for  Southern  Europe  is  also  reproduced:  in  Northern  European  countries, 
we  observe  substantial  volatility  initially  in  careers,  wearing  off  with  increasing  labour  force  experience. 
In  Southern  Europe,  the  pattern  is  one  of  low  volatility  once  employment  has  been  obtained,  even  if 
very  early  after  leaving  the  education  and  training  systems.7 

The  experience  effect  on  unemployment  risks  and  cross-national  differences  therein  which  have  been 
discussed  in  Figure  2 above,  are  also  reproduced  in  the  respective  indicator.  As  discussed  at  length, 
there  is  a division  between  countries  in  terms  of  largely  absent  experience  effects  in  the  set  of 
expected  OLM  countries,  clustering  together  at  the  upper  end  of  the  scale,  compared  to  medium-level 
effects  in  the  other  Northern  European  countries  and  a very  substantial  dependence  of  unemployment 
on  experience  in  Southern  Europe.  With  respect  to  the  issue  of  qualification  effects  on  unemployment, 
there  is  again  mainly  a Northern-Southern  European  divide:  in  all  Northern  European  countries, 
intermediate  education  provides  clear  (and  similar)  advantages  in  terms  of  lower  unemployment  risks 
as  compared  to  compulsory  education  only.  The  relation  is  different  in  Southern  Europe  where  better 
qualified  leavers  regularly  face  higher  unemployment  risks  than  their  lower  qualified  counterparts. 

Turning  to  the  structure  of  occupational  and  status  attainment,  the  results  reported  briefly  above  are 
again  reproduced  in  the  set  of  indicators.  With  respect  to  the  issue  of  lower-skilled  employment,  there 
are  some  indications  of  a particular  pattern  among  expected  OLM  countries.  In  this  set  of  countries, 
recruitment  from  external  markets  typically  includes  a larger  proportion  of  skilled  positions  and 
qualifications  regularly  play  a more  important  role  in  avoiding  lower-skilled  positions.  Moreover,  in 
Austria  and  Germany  at  least,  there  is  little  evidence  of  any  experience  effect  on  the  incidence  of 
lower-skilled  employment,  indicating  a very  clear  difference  of  entry  ports  to  the  market  in  those  two 
countries.  In  part,  this  result  is  also  reflected  in  the  parameters  for  status  attainment.  There  is  some 
evidence  that  the  potential  OLM  countries,  on  average  and  with  some  variation,  differ  from  their 
Northern  European  counterparts,  showing  slightly  weaker  experience  and  somewhat  larger 
qualification  effects  on  status  attainment,  combined  with  less  increase  in  the  qualificational  differential 
over  initial  years  in  the  market.  Moreover,  the  dispersion  of  status  attainment  within  skill  groups  seems 
lower  in  that  set  of  countries  than  in  the  rest  of  Northern  Europe.  Yet,  the  most  pronounced  difference 
is  to  the  Southern  European  countries,  which  all  have  very  strong  qualification  effects  on  status 
attainment  which  generally  change  little  over  initial  careers,  and  an  almost  unchanged  status 
dispersion  with  time  in  the  labour  force.  Attempting  to  make  these  observations  more  systematic,  the 


To  provide  an  example:  the  experience  effect  on  mobility  rates  gives  the  estimated  change  in  that  rate  per 
logged  year  of  labour  force  experience,  i.e.  the  more  negative  the  effects,  the  stronger  the  rate  declines  with 


4.20 


293 


analyses  now  turn  to  the  question  of  whether  it  is  possible  to  discern  from  these  indicators  the 
operation  of  distinct  stratification  systems  in  European  countries  and,  if  so,  in  which  respects  these 
systems  consistently  differ. 


5.  Distinct  Patterns  of  Labour  Market  Entry? 

As  discussed  briefly  in  Section  3 above,  the  similarities  and  differences  in  European  entry  labour 
markets  are  assessed  from  the  outcome  of  country  group  formation  by  means  of  cluster  analyses  and 
the  consistency  of  the  substantive  cross-national  differences  in  stratification  patterns  with  the  paper’s 
theoretical  arguments.  The  next  sub-section  discusses  the  results  achieved  from  that,  while  the 
following  one  contains  the  result  of  subjecting  the  preferred  cluster  solution  to  discriminant  analysis  as 
the  test  for  substantive  differences  in  terms  of  stratification  patterns. 


5.1  Cluster  Analysis  for  the  Set  of  Labour  Market  Indicators 

Figure  7 below  presents  results  from  a cluster  analysis  performed  on  66  European  country-year  cases 
for  the  set  of  structural  indicators  as  detailed  in  Table  2.  Clustering  has  been  carried  out  using  the 
Ward  algorithm  based  on  a squared  Euclidean  distance  matrix  of  z-standardised  transforms  of  the 
labour  market  indicators.  The  figure  presents  both  the  fusion  process  in  terms  of  the  cluster 
dendrogram  and  a set  of  statistics  regularly  reported  for  solution  assessment.  It  is  impossible  to 
discuss  the  choice  of  the  specific  clustering  algorithm  and  the  selected  statistics  shown  here,  but  the 
interested  reader  is  referred  to  e.g.  Bacher  (1994),  Everitt  (1993),  Aldenderfer  and  Blashfield  (1984), 
Kaufman  and  Rousseeuw  (1990)  or  related  literature.  It  may  suffice  here  to  clarify  that  the  Ward 
algorithm  belongs  to  the  broad  class  of  hierarchical  clustering  algorithms  and  specifically  achieves  a 
sequential  fusion  of  least  deviant  cases  or  clusters.  Acknowledging  the  arbitrariness  of  algorithm 
choice,  it  is  a relief  to  be  able  to  note  that  the  substantive  conclusions  from  the  specific  analysis  shown 
appear  reasonably  stable  even  with  some  variation  in  clustering  algorithm  as  well  as  in  analyses 
based  on  country  cases  only. 

From  the  analysis  presented,  the  first  reassurance  relates  to  the  institutionalist  claim  about  the 
importance  of  national  institutional  arrangements,  which  are  thought  of  as  a set  of  stable  context 
factors  underlying  socio-economic  behaviour.  As  there  are  up  to  six  years  available  for  each  country  in 
the  sample,  there  is  some  scope  for  within-country  variation  in  the  set  of  indicators.  Still,  the  fusion 
process  very  clearly  parcels  out  country  clusters  from  country-year  cases  first  and  only  then  proceeds 
to  cluster  country  cases.  This  is  a first  indication  that  annual  variation  in  the  chosen  indicators  is  both 
relatively  less  important  and  occurs  within  national  settings.  A closer  look  at  the  fusion  process  reveals 


experience.  In  turn,  the  most  positive  values  are  close  to  zero,  indicating  almost  absent  experience  effects. 

er|c 


4.21 


294 


Figure  7 Cluster  Analysis  on  Patterns  of  Labour  Market  Entry 


Cluster  Dendrogram 


Rescaled  Distance  Cluster  Combine 
C A S E 0 5 10  15  20 


Label 

Num 

+ 

IT1992 

43 

- + 

IT1993 

44 

- + 

IT1994 

45 

- + 

■ + 

IT1995 

46 

- + 

I 

IT1996 

47 

- + 

I 

IT1997 

48 

- + 

I 

GR1993 

33 

- + 

5 

+ 

GR1995 

35 

- + 

I 

GR1994 

34 

- + — 

- + 

I 

GR1996 

36 

- + 

I 

I 

GR1997 

37 

- + 

I 

I 

GR1992 

32 

- + 

8 

+ — 

- + 

PT1992 

55 

- + 

I 

PT1994 

57 

- + 

I 

PT1997 

60 

- + - - 

- + 

PT1993 

56 

- + 

PT1996 

59 

- + 

PT1995 

58 

- + 

BE1993 

4 

-+  - + 

BE1997 

8 

-+  + • 

FR1996 

30 

-+  I 

I 

FR1997 

31 

-+  - + 

I 

FR1994 

28 

- + 

I 

FR1995 

29 

- + 

I 

FR1993 

27 

- + 

I 

ES1994 

23 

- + 

6 

+ • 

ES1996 

25 

- + 

I 

I 

ES1997 

26 

-+  — 

- + 

I 

I 

ESI  993 

22 

- + 

I 

I 

I 

ESI  995 

24 

- + 

I 

I 

I 

ESI  992 

21 

- + 

1 

I 

I 

UK1996 

65 

- + 

I 

I 

I 

UK 1997 

66 

- + 

9 

+ - - 

- + 

I 

UK1994 

63 

- + 

I 

I 

UK1995 

64 

- + - - 

- + 

I 

I 

UK1993 

62 

- + 

I 

I 

I 

UK1992 

61 

- + 

I 

I 

I 

BE1994 

5 

- + 10  + 

- + 

I 

BE1995 

6 

- + 

I 

I 

BE1996 

7 

- + 

I 

3 +-- 

IE1996 

41 

-+  - - 

- + 

I 

IE1997 

42 

- + 

1 

IE1995 

40 

- + 

I 

I El  993 

38 

- + 

1 

I El  994 

39 

- + 

I 

ATI  995 

1 

- + 

I 

AT1997 

3 

- + - - 

- + 

1 

ATI  9 96 

2 

- + 

+ 

1 

GE1996 

13 

- + 

I 

I 

I 

GE1997 

14 

- + - - 

- + 

I 

I 

GE1994 

11 

- + 

I 

I 

GE1995 

12 

. ~ + 

I 

I 

GE1992 

9 

- + 

1 

1 

GE1993 

10 

- + 

4 

+ — 

NL1993 

50 

- + 

I 

NL1994 

51 

- + 

1 

NL1992 

49 

- + 

I 

NL1995 

52 

-+  - - 

- + 

I 

NL1996 

53 

- + 

I 

I 

NL1997 

54 

- + 

7 

+ - - 

DK1993 

16 

- + 

I 

DK1994 

17 

- + - + 

I 

DK1997 

20 

- + + 

-- 

- + 

DK1995 

18 

-+  I 

DK1996 

19 

- + - + 

DK1992 

15 

- + 

Cluster  Solution  Statistics 

(only  last  10  steps) 


25 

K 

Vi 

SSQin 

?a 

PRE 

10 

118.5 

159.4 

0.81 

0.15 

9 

138.1 

187.5 

0.78 

0.15 

8 

167.6 

221.9 

0.74 

0.17 

7 

198.2 

266.7 

0.69 

0.15 

6 

231.8 

314.1 

0.63 

0.15 

1 

I 

5 

279.2 

371.4 

0.57 

0.17 

I 

I 

4 

334.1 

446.3 

0.48 

0.16 

I 

3 

389.6 

528.9 

0.38 

0.17 

I 

I 

2 

502.1 

640.7 

0.25 

0.25 

I 

I 

1 

715.0 

859.6 

- 

- 

I 

I ’ 
I 

Mojena 

Mojena 

I 

K 

1"  Beale 

Sign. 

1 

II 

I 

I 

10 

3.23 

0.08 

4.67 

2.57 

I 

9 

3.42 

0.07 

5.19 

3.01 

2 I 

I 

8 

3.84 

0.05 

5.51 

3.24 

I 

I 

7 

3.44 

0.07 

5.64 

3.35 

I 

I 

6 

3.59 

0.06 

5.91 

3.61 

I 

5 

4.03 

0.05 

6.12 

3.78 

I 

I 

4 

3.77 

0.06 

6.08 

3.76 

I 

I 

3 

4.37 

0.04 

6.71 

4.35 

I 

2 

7.18 

0.01 

8.27 

5.50 

I 

I 

1 

- 

- 

- 

- 

i 

i 

+ Inverse  Scree  Diagram 


Notes:  Clustering  is  carried  out  by  applying  the  Ward  algorithm  using  a squared  Euclidean  distance  matrix 

based  on  z-standardised  transforms  of  labour  market  indicators  (cf.  Table  2 and  Figure  6). 

Source:  European  Community  Labour  Force  Surveys,  1992-1997 


4.22 


Fisioa  level 


that  actually  more  than  85%  of  the  variation  in  the  set  of  indicators  as  measured  by  n2  is  between 
countries,  while  less  than  15%  is  due  to  within-country  annual  variation. 

What  then  is  the  substantive  content  of  the  country  clustering  achieved  - and  how  well  do  countries 
cluster?  Judged  from  the  dendrogram  depiction  of  the  fusion  process  and  the  fit  statistics,  especially 
the  most  conservative  FBeaie-test  applied  here,  a solution  distinguishing  three  clearly  separated  country 
clusters  appears  most  appropriate.  The  country  clusters  distinguished  thus  are  (1)  a cluster  of 
Southern  European  countries  comprising  Italy,  Greece  and  Portugal,  (2)  a cluster  of  North-Western 
European  countries  including  Belgium,  France,  Ireland,  the  United  Kingdom,  but  also  Spain,  and  (3)  a 
final  cluster  consisting  of  Austria,  Denmark,  Germany  and  the  Netherlands.  Apparently,  this  result  has 
two  immediate  implications:  first,  the  distinction  within  the  group  of  Northern  European  countries  is 
apparently  well  in  line  with  both  institutionalist  arguments  and  current  research  reviewed  extensively  in 
earlier  sections.  Those  countries  where  education  and  training  systems  are  strongly  vocationally 
orientated  are  clearly  separated  from  those  countries  that  do  not  have  such  training  arrangements;  in 
this  sense,  there  is  some  support  for  arguments  about  the  existence  of  distinct  stratification  systems  - 
the  precise  nature  of  which  will  be  assessed  below  - and  their  close  relation  to  institutional 
arrangements  of  linking  education  and  training  systems  to  labour  markets.  But  apart  from  that  and 
unexpected  from  the  perspective  of  a theory  linking  stratification  systems  to  institutional  arrangements 
in  training  systems,  the  stratification  patterns  observed  for  - broadly  speaking  - the  Southern 
European  countries  also  stand  out  distinctly.  Thus,  while  the  popular  dichotomy  of  stratification 
systems  presumably  taps  an  important  aspect  within  Northern  European  labour  markets,  it  is  far  from 
clear  how  the  Southern  countries  fit  into  that  one-dimensional  framework.  Rather,  it  seems  that 
alternative  institutional  arguments  have  to  be  supplemented  in  order  to  provide  a satisfactory  account 
of  Southern  patterns  of  market  entry.  Some  suggestions  regarding  this  issue  will  be  developed  in  the 
concluding  section. 

Returning  for  a moment  to  the  results,  some  caveats  about  the  approriateness  of  the  cluster  solution 
have  to  be  added  here.  On  the  one  hand,  the  simple  three  cluster  solution  chosen  for  further 
investigation  seems  remarkably  powerful  in  terms  of  “explained"  variation  as  it  captures  roughly  40% 
of  the  overall  variation  in  the  set  of  indicators.  This  gives  yet  another  indication  of  the  extent  to  which 
major  differences  in  labour  market  entry  are  effectively  located  between  broad  sets  of  countries.  On 
the  other  hand,  the  question  naturally  arises  whether  the  chosen  fusion  point  is  the  most  natural  to 
stop  at.  At  least,  the  less  conservative  Mojena  I/ll  tests  would  suggest  the  country  level  clustering  or  a 
nine  cluster  solution  as  more  appropriate;  moreover,  the  PRE  error  reduction  brought  about  by  lower 
levels  of  aggregation  is  still  substantial  and  even  the  F-test  for  marginal  improvement  is  hovering 
around  p-levels  of  .05.  So  a more  cautious  interpretation  would  clearly  be  that  there  are  important 
national  differences  within  each  country  cluster  distinguished,  which  then  simply  cannot  be  addressed 
from  the  very  generalist  framework  adopted  here.  On  the  other  hand,  one  still  has  to  acknowledge  that 
this  occurs  against  the  background  of  a clear  distinction  of  three  sets  of  countries,  so  that  further 
country  differences  appear  minor  compared  to  those  differences  in  stratification  systems  explored 
here.  Still,  the  immediately  adjacent  four-cluster  solution  may  deserve  special  attention  in  further 


4.23 


296 


research  for  a number  of  reasons.  This  solution  offers  an  additional  split  within  the  OLM  countries 
between  Austria  and  Germany  on  the  one  hand  and  Denmark  and  the  Netherlands  on  the  other.  The 
fact  that  this  is  the  “next  most  important  split”  in  the  data,  which  corresponds  very  well  with  the 
difference  between  the  countries  in  terms  of  school-based  versus  apprenticeship-based  provision  of 
vocational  training,  may  suggest  that  these  countries'  institutional  arrangements  are  each  located  at  a 
particular  threshold  of  “OLM-likeness"  in  the  stratification  system.  Of  course,  this  issue  would  best  be 
pursued  in  more  detailed  case  studies  of  these  countries.  Naturally,  this  applies  to  other  established 
contrasts  as  well,  and  some  comments  on  this  will  follow  in  the  concluding  section. 

Having  identified  the  above  three-cluster  solution  as  the  main  result  of  the  clustering  step,  Table  3 
finally  presents  results  from  a small-scale  sensitivity  analysis  for  that  solution,  based  on  the  deletion  of 
single  indicators  from  the  calculation  of  the  distance  matrix.  After  all,  the  substantive  cluster  solution 
preferred  also  exhibits  a sensible  degree  of  stability  in  that  exercise.  As  judged  from  the  results  given 
in  Table  3,  no  single  aspect  of  the  stratification  of  early  careers  is  of  decisive  importance  for  arriving  at 
the  solution  discussed  here.  Rather,  the  clustering  outcome  presented  here  seems  to  follow  from  the 
simultaneous  consideration  of  the  full  range  of  indicators;  deleting  single  indicators  regularly  induced 
little  change  in  the  results.  Of  all  variables  under  study,  the  issue  of  unemployment  deserves  special 
attention  as  probably  the  most  influential  aspect  in  the  analyses:  removing  unemployment  from  the 
analyses  actually  leads  to  a major  reallocation  of  the  countries,  namely  an  allocation  of  Denmark  and 
the  Netherlands  together  with  the  group  of  Northern  ILM-type  countries  rather  than  with  Austria  and 
Germany,  providing  further  evidence  of  some  heterogeneity  within  the  small  set  of  OLM  systems. 


Table  3 Sensitivity  Analysis  for  Cluster  Analysis  Results 


Variables  included 

3-Cluster  Solution 

4-Cluster  Solution 

Base:  Full  model 

(GR,IT,PT)  - (BE.ES.FR, IE, UK)  - 
(AT.GE.DK.NL) 

(GR.IT.PT)  - (BE.ES.FR, IE, UK)  - 
(AT.GE)  - (DK.NL) 

(1)  B - labour  mobility 

(GR,IT,PT)  - (BE.DK.ES, FR, IE, UK) 
- (AT.GE.NL) 

(IT)  - (GR.PT)  - 

(BE.DK.ES.FR.IE.UK)  - (AT.GE.NL) 

(2)  B - lower-skill  bias  in  recruiting 

(GR,JT,PT)  - (BE.ES.FR, IE, UK)  - 
(AT.GE.DK.NL) 

(IT)  - (GR.PT)  - (BE.ES.FR, IE, UK) 
-(AT.GE.DK.NL) 

(3)  B - lower-skilled  employment 

(GR,IT,PT)  - (AT, BE.ES.FR, IE, UK) 
- (GE.DK.NL) 

(IT)  - (GR.PT)  - 

(AT, BE.ES.FR, IE, UK)  - (GE.DK.NL) 

(4)  B - unemployment 

(BE.DK.ES.FR.IE.NL.UK)  - 
(GR.PT)  - (AT,GE,IT) 

(GR.PT)  - (BE.ES.FR, IE, UK)  - 
(DK.NL)  - (AT.GE, IT) 

(5)  B - status  attainment 

(GR.IT.PT)  - (BE.ES.FR, IE, UK)  - 
(AT.GE.DK.NL) 

(GR.IT.PT)  - (BE.ES.FR, IE, UK)  - 
(AT.GE) -(DK.NL) 

(6)  B - status  attainment  dispersion 

(GR,IT,PT)  - (BE.ES.FR, IE, UK)  - 
(AT.GE.DK.NL) 

(GR.IT.PT)  - (BE.ES.FR, IE, UK)  - 
(AT.GE) -(DK.NL) 

Notes:  Clustering  is  carried  out  by  applying  the  Ward  algorithm  using  a squared  Euclidean  distance  matrix 
based  on  z-standardised  transforms  of  labour  market  indicators  (cf.  Table  2 and  Figure  6);  reference  full 
model  is  the  one  detailed  in  Figure  7. 

Source:  European  Community  Labour  Force  Surveys,  1992-1997. 


O 

ERIC 


4.24 


297 


5.2  Distinctive  Features  of  Market  Entry  in  Different  Stratification  Systems 

As  a final  step  in  the  analysis,  the  substantive  differences  between  stratification  systems  as  identified 
from  the  three  cluster  solution  singled  out  above  are  of  primary  interest.  Table  4 below  presents  the 
structure  matrix  and  related  statistics  from  a discriminant  analysis  of  the  three  sets  of  countries  in 
terms  of  the  set  of  indicators. 

The  main  outcome  of  this  final  analysis  is  the  extraction  of  two  discriminant  functions  summarising 
distinctive  features  of  the  three  stratification  systems  distinguished.  Among  these,  the  first  and  more 
powerful  one  distinguishes  the  Southern  European  from  the  two  Northern  European  country  clusters: 
the  three  Southern  European  countries  (excluding  Spain)  are  located  in  the  positive  area  of  the 
function,  while  all  Northern  European  countries  (including  Spain)  tend  towards  the  negative  end  of  the 
scale.  According  to  the  correlations  between  indicators  and  canonical  discriminant  functions  as  given 
in  the  structure  matrix,  there  are  apparently  three  main  factors  distinguishing  the  Southern  European 
countries  from  the  remaining  EU  countries:  (1)  low  rates  of  mobility  in  early  career,8  (2)  low 
qualification  effects  on  market  exclusion,  especially  unemployment,  and  (3)  high  qualification  effects 
on  attainment.  The  magnitudes  of  the  correlations  indicate  that  differences  in  mobility  rates  and  the 
qualificationa!  stratification  of  unemployment  are  the  most  important  factors  differentiating  Southern 
European  from  Northern  European  countries,  although  the  differences  in  skill  differentials  in  status 
attainment  are  certainly  also  pronounced. 

The  second  discriminant  function  then  mainly  separates  the  two  Northern  European  country  clusters 
(including  Spain),  broadly  consistent  with  the  expectation  of  a dichotomy  of  ILM-type  versus  OLM-type 
stratification  systems.  The  differentiating  aspects  here  are  (1)  lower  experience  effects  on 
unemployment,  (2)  a smaller  bias  towards  lower  skilled  jobs  in  external  recruitment,  (3)  lower 
experience  effects  on  attainment  dispersion,  (4)  stronger  qualification  effects  on  the  incidence  of 
unemployment,  and  (5)  lower  experience  effects  on  the  incidence  of  lower  skilled  employment  in  the 
group  formed  by  Austria,  Denmark,  Germany,  and  the  Netherlands  as  compared  to  the  other  Northern 
European  countries.  Comparing  the  magnitude  of  correlations  again,  the  first  two  factors  appear  the 
most  relevant  in  drawing  the  line  between  the  two  groups  of  countries.  That  is,  the  group  of  OLM-type 
systems  is  mainly  distinguished  from  its  ILM-type  counterparts  in  terms  of  unemployment  not  being 
concentrated  on  market  entrants  (but  rather  on  the  lowest  qualified)  and  a stronger  tendency  also  to 
hire  into  skilled  jobs  from  external  markets.  Apart  from  that,  the  dispersion  of  status  attainment  with 
time  in  the  labour  force  is  lower  and  young  people,  finally,  also  move  less  out  of  lower  skilled 
employment  with  time  in  the  market,  mostly  because  the  incidence  of  such  employment  among 


8 Consider  again  the  different  types  of  experience  effects  as  an  example:  the  experience  effects  on 
unemployment  or  mobility  give  the  estimated  change  in  unemployment  or  mobility  rates  per  logged  year  of 
labour  force  experience,  i.e.  the  more  negative  the  effects,  the  more  strongly  the  respective  rates  decline  with 
experience.  In  turn,  the  most  positive  values  are  close  to  zero,  indicating  almost  absent  experience  effects. 
The  respective  entries  in  the  structure  matrix  thus  are  to  be  read  in  the  sense  that  the  more  positive,  i.e. 

O 

4.25 


Table  4 Discriminant  Analysis  of  Country  Cluster  Solution:  Structure  Matrix 


Discriminant  Function 

Indicator  1 2 


Unemployment:  Qualification  Effect 

0.390 

0.181 

Labour  Force  Mobility:  Experience  Effect 

0.370 

-0.081 

Labour  Force  Mobility:  Rate 

-0.335 

0.087 

Status  Attainment:  Qualification  Effect 

0.258 

0.038 

Lower-Skilled  Employment:  Qualification  Effect 

0.159 

-0.139 

Unemployment:  Experience  Effect 

-0.149 

0.340 

External  Mobility:  Lower  Skill  Bias 

0.042 

-0.325 

Status  Attainment  Dispersion:  Experience  Effect 

-0.101 

-0.186 

Lower-Skilled  Employment:  Experience  Effect 

0.089 

0.153 

Status  Attainment:  Experience  Effect 

0.024 

-0.099 

Status  Attainment:  Experience-Qualification  Interaction 

-0.047 

0.047 

Eigenvalue 

18.807 

9.528 

Proportion  of  Variance 

66.4  % 

33.6  % 

Canonical  Correlation 

0.974 

0.951 

Discriminant  Functions  at  Group  Centroids 


1:  AT  GE  DKNL 

-3.262 

3.755 

2:  BE  ES  FR  IRE  UK 

-2.041 

-3.321 

3:  IT  GR  PT 

6.867 

0.601 

Notes:  Cluster  solution  analysed  is  the  three-cluster  solution  singled  out  of  the  analysis  detailed  in  Figure  7. 

Source:  European  Community  Labour  Force  Survey  1992-1997. 


intermediate  skills  is  lower  from  the  outset  of  careers.  After  all,  this  empirical  evidence  on  the 
substantive  features  differentiating  both  types  of  Northern  European  countries  does  appear  very  much 
in  line  with  the  expectations  derived  from  the  institutional  reasoning  about  ILM  and  OLM  types  of 
systems  discussed  at  length  in  the  theoretical  sections  above.  It  clearly  is  the  case  that  these  sets  of 
countries  differ  in  the  relative  reliance  of  attainment  processes  on  either  educational  skills  or  labour 
force  experience.  As  such,  it  seems  reasonable  to  think  of  this  contrast  as  consequence  of  the  specific 
institutional  mechanisms  of  allocation  as  suggested  by  Marsden’s  framework  or  related  works.  Still,  it 
has  to  be  recognised  that  the  country  contrast  is  driven  by  differences  in  terms  of  both  employment 
and  unemployment  patterns,  probably  even  more  strongly  by  differences  in  the  latter.  Such  a result  is 
not  fully  captured  in  current  institutional  reasoning  centred  around  the  association  between  education 
and  occupational  outcomes,  though  the  theoretical  part  of  this  paper  already  attempted  to  argue  about 
consistent  differences  between  ILM  and  OLM  systems  in  both  market  exclusion  and  market  attainment 
in  early  careers.  Apart  from  that,  the  most  intriguing  weakness  of  the  currently  used  dichotomy  of 
stratification  systems  is  that  - although  it  has  been  possible  to  identify  a distinct  Southern  European 
cluster  based  on  critical  parameters  suggested  by  current  institutional  reasoning  - it  offers  little 
substantive  explanation  for  this  pattern.  More  comments  on  this  follow  below. 


absent,  experience  effects  on  mobility  and  unemployment,  the  higher  the  discriminant  score  on  function  1 or  2 
respectively. 


4.26 


299 


6.  Conclusions 


What  then  has  been  learned  about  entry  labour  markets  in  European  Union  countries  and  the  relation 
of  cross-national  differences  to  differences  in.  the  institutional  linkage  between  training  systems  and 
labour  markets?  The  concluding  section  of  this  paper,  attempts  both  to  summarise  the  empirical 
results  on  labour  market  entry  in  Europe  and  to  draw  a set  of  conclusions,  touching  upon  issues  of 
research  design  and  the  possibility  of  empirical  assessments  of  institutionalist  arguments,  the  quality 
of  information  in  the  database  employed  and  which  information  it  fails  to  provide,  as  well  as  the 
analytical  power  of  current  theorising  in  comparative  stratification  research  about  the  existence  and 
effects  of  a dichotomy  of  stratification  systems,  linked  to  the  occupational  specificity  of  education  and 
training  systems. 

At  first  glance,  the  current  paper  simply  provided  a rich  set  of  empirical  results  on  different  aspects  of 
the  process  of  labour  market  integration  in  the  countries  of  the  European  Union.  And  although  some 
broad  similarities  do  appear  from  the  analyses,  the  major  outcome  is  to  establish  consistent 
differences  between  sets  of  European  countries  in  terms  of  crucial  features  of  labour  market  entry 
processes.  Indeed,  here  the  analytical  value  of  the  notion  of  stratification  systems  is  self-evident: 
identifying  relevant  sources  of  cross-national  differences  clearly  enough  to  allow  for  operationalisation 
and  empirical  measurement  of  core  concepts  is  well  on  the  way  to  understanding  the  impact  of 
different  institutional  arrangements  in  these  societies.  Allocation  mechanisms  in  different  systems  vary 
in  their  relative  reliance  on  either  educational  skills  or  experience  and  mobility  - and  in  consequence 
lead  to  different  patterns  of  labour  market  entry. 

Summarising  the  empirical  results  of  this  study,  there  is  indeed  substantial  support  for  an  institutional 
account  of  cross-national  differences  in  labour  market  entry  patterns.  On  the  one  hand,  institutionalist 
reasoning  is  supported  by  the  general  result  that  most  cross-national  variation  is  actually  variation  in 
national  contexts,  irreducible  to  short-term  variation  of  whatever  origin.  Moreover,  the  notion  of 
stratification  systems  gains  considerable  appeal  from  noting  how  much  of  the  cross-country  variation 
in  labour  markets  is  effectively  captured  within  a small  subset  of  country  clusters  as  distinguished 
here.  Entry  labour  markets  are  clearly  distinct  in  terms  of  features  such  as  the  extent  of  recruitment 
into  lower-level  entry  port  jobs,  the  scope  for  upward  occupational  mobility  or  the  stratification  of 
market  exclusion.  Specifically,  the  contrast  among  Northern  European  countries  between  an  ILM 
system  group  formed  by  the  United  Kingdom,  France,  Ireland  and  Belgium,  and  a set  of  OLM  systems 
operating  in  Austria,  Denmark,  Germany,  and  the  Netherlands  differing  precisely  in  the  relative 
importance  of  education  or  experience  in  allocation  processes  appears  closely  in  line  with  current 
institutional  arguments.  There  is  a larger  role  for  experience  effects  and  mobility  in  channelling  the 
flow  of  individuals  into  positions  in  the  context  of  institutional  arrangements  provided  in  the  set  of  these 
ILM  countries.  To  put  it  slightly  differently,  the  less  allocation  mechanisms  in  stratification  systems  rest 
on  experience  and  mobility  criteria,  the  less  early  labour  market  careers  differ  from  later  ones  and  the 
less  “problematic”  labour  market  entry  appears.  Given  the  close  coincidence  of  empirically  identified 
systems  with  types  of  institutional  arrangements  in  education  and  training  systems,  it  is  quite  plausible 


ERIC 


4.27 


300 


that  the  interlinkage  of  education  and  training  systems  and  labour  markets  is  a key  institutional  factor 
in  transforming  stratification  systems. 

Of  course,  much  of  this  result  rests  on  the  specific  context  and  empirical  approach  of  the  study. 
Actually,  this  study  claims  no  more  than  to  provide  a serious  proposal  for  identifying  the  impact  of 
institutional  arrangements  empirically,  extending  Marsden's  earlier  attempts  (1990;  Eyraud  et  al. 
1990).  Naturally,  concrete  operationalisations  owed  as  much  to  database  content  and  limitations  as  to 
the  original  theoretical  concepts.  As  such,  the  major  limitation  of  using  the  European  Community 
Labour  Force  Surveys  has  to  be  seen  in  their  fully  cross-sectional  design  which  prevents  one  from 
exploring  mobility  structures  in  any  detail.  An  extended  replication  of  at  least  parts  of  the  analyses  with 
longitudinal  data  would  certainly  be  warranted  for  further  validation  of  the  conclusions  drawn  here. 
Moreover,  replication  of  even  the  present  research  setup  on  a larger  set  of  countries,  including  non- 
European  Union  ones,  could  contribute  to  a further  and  stricter  test  of  the  general  argument.  Still,  both 
the  substantive  and  analytical  results  from  this  study  potentially  provide  some  direction  for  further 
research. 

Based  on  the  analyses  conducted  here,  there  are  at  least  two  obvious  points  of  departure  for  further 
inquiry:  first,  investigating  further  cross-national  differences  within  the  major  types  of  stratification 
systems  more  closely,  and  second,  attempting  to  incorporate  the  existence  of  a distinct  system  type  of 
Southern  European  countries  into  the  theoretical  toolkit  of  comparative  stratification  research.  Turning 
to  the  former  issue  first,  it  is  obvious  that  the  amount  of  within-cluster  variation  points  to  the  fact  that 
although  the  ILM-/OLM-system  contrast  taps  an  important  aspect  of  cross-national  differences  in 
labour  market  entry,  it  is  far  from  being  the  only  relevant  one.  National  variations  within  the  broader 
configurations  and  the  source  of  deviating  features  certainly  form  a worthwhile  object  for  further 
empirical  and  institutional  analyses.  As  suggested  in  Section  5,  the  contrast  between  OLM 
arrangements  in  Austria  and  Germany  versus  those  in  Denmark  or  the  Netherlands  is  an  especially 
interesting  case,  as  the  results  could  be  read  as  pointing  to  the  existence  of  institutional  thresholds  of 
“OLM-likeness"  of  systems.  Alternatively,  one  might  argue  that  labour  market  entry  in  Denmark  and 
the  Netherlands  currently  appears  similar  to  patterns  found  for  Austria  and  Germany  only  because  of 
the  presently  favourable  aggregate  labour  market  situation  in  these  countries.  If  so,  then  only 
extended  historical  comparisons  will  yield  more  definite  answers.  As  this  single  example  should 
suggest,  additional  and  well-directed  in-depth  comparative  case  studies  for  crucial  country  cases  may 
well  be  expected  to  provide  future  fine  tuning  of  institutional  explanations  of  labour  market  entry. 

Second,  it  seems  equally  relevant  to  theoretically  acknowledge  the  operation  of  a distinct  Southern 
European  pattern  of  stratification  in  both  the  analysis  of  labour  market  entry  and  comparative 
stratification  research  in  general  (cf.  also  Roberts  1999).  It  is  apparent  that  Southern  European 
countries  show  a specific  mixture  of  structural  features  of  early  labour  market  careers:  here  both 
strong  qualification  and  strong  experience  effects  occur,  in  conjunction  with  very  high  unemployment 
risks  at  the  outset  of  careers,  yet  rather  little  volatility  once  initial  employment  has  been  secured.  That 
is,  the  Southern  European  patterns  appears  to  mix  elements  present  in  ILM-type  systems  - as  in  the 


0 

ERIC 


4.28 


role  of  experience  effects  on  lower  skilled  employment  - and  OLM  arrangements  - where  larger 
qualificational  differentials  in  rewards  appear  - with  their  specific  peculiarities  such  as  an  absence  of 
qualificational  stratification  of  unemployment  and  low  mobility  even  at  market  entry.  Some  of  these 
more  ILM-type  features  were  to  be  expected  considering  the  institutional  arrangements  in  Southern 
European  education  and  training  systems.  Still,  major  parts  of  the  findings  do  not  fit  easily  into  a 
dichotomous  contrast  of  ILM  versus  OLM  systems,  as  derived  from  considering  variation  in  the 
institutional  structure  of  education  and  training  systems  as  the  main  variable  of  interest. 

A superficial  glance  at  alternative  literature  seems  to  suggest  the  importance  of  at  least  two  additional 
institutional  complexes,  which  should  make  it  possible  to  integrate  the  “Mediterranean”  model  into  an 
institutional  argument,  namely  labour  market  legislation  and  the  role  of  the  family.  Interestingly,  there 
seems  to  be  initial  support  for  both  strands  of  the  argument:  first,  the  “deviant”  case  of  Spain  could 
effectively  stem  from  a deliberate  political  attempt  to  alleviate  the  perceived  obstacle  of  strict  labour 
legislation  and  to  make  youth  labour  markets  more  flexible  by  introducing  fixed-term  contracts  - and 
the  excessive  use  of  these  types  of  arrangements  afterwards  transforming  the  stratification  system 
into  the  ILM  model  (cf.  Bentolila  and  Dolado  1994  for  an  overview  of  the  changes  that  have  occurred): 
On  the  other  hand,  it  is  regularly  reported  for  Southern  European  countries  that  extensive  family 
support  enables  young  people  to  wait  until  adequate  employment  can  be  attained  (Bernardi  et  al. 
1999),  providing  a consistent  account  for  the  observed  combination  of  strong  qualification  effects  on 
occupational  attainment  in  conjunction  with  strong  experience  effects  on  market  exclusion.  In 
combination  with  still  effective  strict  labour  legislation  reducing  volatility  and  mobility  in  the  three 
Southern  European  countries  clustered  together  here,  both  strands  could  yield  a consistent  argument 
on  the  institutional  foundations  of  this  system.  Naturally,  only  future  research  will  be  able  to  provide 
adequate  answers. 


O 

ERIC 


4.29 


302 


References 


Aldenderfer,  Mark  S.  and  Roger  K.  Blashfield,  1984:  Cluster  Analysis.  Beverly  Hills:  Sage. 

Allmendinger,  Jutta,  1989:  Educational  systems  and  labour  market  outcomes.  European  Sociological 
Review  5,  231-250. 

Ashton,  David  N.,  1988:  Sources  of  Variation  in  Labour  Market  Segmentation:  A Comparison  of  Youth 
Labour  Markets  in  Canada  and  Britain.  Work,  Employment  and  Society  2,  1-24. 

Ashton,  David  N.,  Malcolm  Maguire  and  Mark  Spilsbury,  1990:  Restructuring  the  Labour  Market.  The 
Implications  for  Youth.  Houndmills:  Macmillan. 

Ashton,  David  N.,  1994:  Understanding  change  in  youth  labour  markets:  a conceptual  framework.  In 
CEDEFOP:  The  determinants  of  transitions  in  youth.  Berlin:  CEDEFOP. 

Ashton,  David  N.,  1997:  Labour  market  approaches  to  the  study  of  the  relationship  between  education 
and  (un)employment  in  the  United  Kingdom.  Pp.  143-159  in:  Jobert  et  al.,  1997. 

Bacher,  Johann,  1994:  Clusteranalyse:  anwendungsorientierte  Einfuhrung.  Munchen:  Oldenbourg. 

Bentolila  and  Dolado,  1994:  Labour  flexibility  and  wages:  lessons  from  Spain.  Economic  Policy  18:  53- 
99. 

Bernardi,  Fabrizio,  Richard  Layte,  Antonio  Schizzerotto  and  Sheila  Jacobs,  1999:  Who  exits 
unemployment?  Institutional  features,  individual  characteristics  and  chances  of  getting  a job.  A 
comparison  of  Britain  and  Italy.  Presentation  to  the  ESA  4th  European  Conference  of 
Sociology.  Amsterdam:  mimeo. 

Brauns,  Hildegard,  Walter  Muller,  and  Susanne  Steinmann,  1998:  Education  and  Social  Class:  A 
Comparative  Study  on  the  Consequences  of  Educational  Expansion  in  Germany,  France,  the 
UK  and  Hungary.  Pp.  75-130  in:  Network  on  Transitions  in  Youth  (Eds.),  Transitions  in  Youth: 
Combating  Exclusion.  Dublin:  ESRI. 

Doeringer,  Peter  B.,  and  Michael  J.  Piore,  1971:  Internal  labor  markets  and  manpower  analysis. 
Lexington:  Heath. 

Eurostat,  1996:  The  European  Union  labour  force  survey.  Methods  and  definitions  1996.  Luxembourg: 
Eurostat. 

Everitt,  Brian  S,,  1993:  Cluster  Analysis.  3rd  edition.  London:  Arnold. 

Eyraud,  Frangois,  David  Marsden,  and  Jean-Jacques  Silvestre,  1990:  Occupational  and  internal 
labour  markets  in  Britain  and  France.  International  Labour  Review  129,  501-517. 

Ganzeboom,  Harry  B.  G.,  Paul  M.  De  Graaf,  and  Donald  Treiman,  1992:  A Standard  International 
Socio-Economic  Index  of  Occupational  Status.  Social  Science  Research  21,  1-56. 

Ganzeboom,  Harry  B.  G.,  and  Donald  Treiman,  1996:  Internationally  Comparable  Measures  of 
Occupational  Status  for  the  1988  International  Standard  Classification  of  Occupations.  Social 
Science  Research  25,  201-239. 

Garonna,  Paolo  and  Paul  Ryan,  1989:  Le  travail  des  jeunes,  les  relations  professionnelles  et  les 
politiques  sociales  dans  les  economies  avancbes.  Formation  Emploi  n°25,  janvier-mars. 

Hall,  Peter  A.,  and  David  Soskice,  1998:  Varieties  of  Capitalism:  The  Institutional  Foundations  of 
Comparative  Advantage.  Berlin:  mimeo. 


«'*V  ■ 


4.30 


303 


Hannan,  Damian  F.,  David  Raffe,  and  Emer  Smyth,  1997:  Cross-National  Research  on  School  to 
Work  Transitions:  An  Analytical  Framework.  Pp.  409-442  in:  Patrick  Werquin,  Richard  Breen, 
and  Jordi  Planas  (Eds.),  Youth  Transitions  in  Europe:  Theories  and  Evidence.  Marseille: 
CEREQ. 

Hannan,  Damian  F.,  et  al.,  1999:  A Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe  (CATEWE):  A Conceptual  Framework.  ESRI  Working  Paper  No.  138(a).  Dublin:  The 
Economic  and  Social  Research  Institute. 

International  Labour  Organisation,  1990:  Statistical  Sources  and  Methods.  Volume  3:  Economically 
Active  Population,  Employment,  Unemployment  and  Hours  of  Work  (Household  Surveys).  2nd 
edition.  Geneva:  ILO. 

Jobert,  Annette,  Catherine  Marry,  Lucie  Tanguy,  and  Helen  Rainbird  (Eds.),  1997:  Education  and 
Work  in  Great  Britain,  Germany  and  Italy.  London:  Routledge. 

Jovanovic,  Boyan,  1979:  Job  Matching  and  the  Theory  of  Turnover.  Journal  of  Political  Economy  87, 
972-990. 

Kaufman,  Leonard  and  Peter  J.  Rousseeuw,  1990:  Finding  Groups  in  Data.  An  Introduction  to  Cluster 
Analysis.  New  York:  Wiley. 

Kerckhoff,  Alan  C.,  1995:  Institutional  Arrangements  and  Stratification  Processes  in  Industrial 
Societies.  Annual  Review  of  Sociology  15,  323-347. 

Kerckhoff,  Alan  C.,  1996:  Building  Conceptual  and  Empirical  Bridges  between  Studies  of  Educational 
and  Labour  Force  Careers.  Pp.  37-56  in:  Alan  C.  Kerckhoff  (ed.),  Generating  Social 
Stratification:  Toward  a New  Research  Agenda.  Boulder:  Westview  Press. 

Kerckhoff,  Alan  C.,  forthcoming:  Transition  from  School  to  Work  in  Comparative  Perspective.  To 
appear  in  Maureen  T.  Hallinan  (ed.),  Handbook  of  Sociology  of  Education,  mimeo. 

LeGrand,  Carl,  and  Michael  Tcihlin,  1998:  Job  Mobility  and  Earnings  Growth.  SOFI  Meddelanden  8/98. 
Stockholm:  Institutet  for  Social  Forskning. 

Marsden,  David,  1986:  The  End  of  Economic  Man?  Custom  and  Competition  in  Labour  Markets. 
Brighton:  Wheatsheaf.  . 

Marsden,  David,  1990:  Institutions  and  Labour  Mobility:  Occupational  and  Internal  Labour  Markets  in 
Britain,  France,  Italy  and  West  Germany.  Pp.  414-438  in:  Renato  Brunetta  and  Carlo 
Dell'Aringa  (Eds.),  Labour  Relations  and  Economic  Performance.  Houndmills:  Macmillan. 

Marsden,  David,  1992:  Segmentation  des  marches  du  travail  et  chomage  de  longue  duree  Pp.  113- 
122  in:  Patricia  Bouillaguet  et  Christophe  Guitton  (Eds.),  Le  chomage  de  longue  duree: 
comprendre,  agir,  evaluer.  Collection  “TEN”,  Paris,  Syros-Alternatives. 

Marsden,  David,  1993:Existera-t-il  un  grand  marche  du  travail  europeen?  Formation  Emploi  n°43, 
juillet-septembre. 

Marsden,  David,  1997:  The  approaches  of  economists  to  the  relationship  between  training  and  work. 
Pp.  1 92-207  in:  Jobert  et  al.,  1 997. 

Marsden,  David,  and  Paul  Ryan,  1995:  Work,  labour  markets  and  vocational  preparation:  Anglo- 
German  comparisons  of  training  in  intermediate  skills.  Pp.  67-79  in:  Leslie  Bash  and  Andy 
Green  (Eds.),  Youth,  Education  and  Work.  London:  Kogan  Page. 

Maurice,  Marc,  Frangois  Sellier  and  Jean-Jacques  Silvestre,  1986:  The  social  foundations  of  industrial 
power:  a comparison  of  France  and  Germany.  Cambridge:  MIT  Press. 


4.31 


304 


Muller,  Walter,  and  Yossi  Shavit,  1998:  The  Institutional  Embeddedness  of  the  Stratification  Process: 
A Comparative  Study  of  Qualifications  and  Occupations  in  Thirteen  Countries.  Pp.  1-48  in: 
Yossi  Shavit  and  Walter  Muller  (Eds.),  From  School  to  Work.  A Comparative  Study  of 
Educational  Qualifications  and  Occupational  Destinations.  Oxford:  Clarendon  Press. 

OECD,  1997:  Education  at  a Glance.  OECD  Indicators  1997.  Paris:  OECD. 

Roberts,  Ken  et  al.,  1999:  The  Young  Unemployed  and  the  Mediterraneanisation  of  Youth  Transition 
in  East-Central  Europe.  Paper  presented  to  the  1999  Annual  Meeting  of  the  European 
Research  Network  on  Transitions  in  Youth.  Oslo:  mimeo. 

Rosenbaum,  James  E.,  Takehiko  Kariya,  Rick  Settersten,  and  Tony  Maier,  1990:  Market  and  Network 
Theories  of  the  Transition  from  High  School  to  Work:  Their  Application  to  Industrialized 
Societies.  Annual  Review  of  Sociology  16,  263-299. 

Sengenberger,  Werner,  1987:  Struktur  und  Funktionsweise  von  Arbeitsmcirkten.  Die  Bundesrepublik 
Deutschland  im  internationalen  Vergleich.  Frankfurt,  New  York:  Campus. 

Shavit,  Yossi,  and  Walter  Muller,  forthcoming:  Vocational  Secondary  Education,  Tracking  and  Social 
Stratification.  To  appear  in  Maureen  T.  Hallinan  (ed.),  Handbook  of  Sociology  of  Education, 
mimeo. 

Soskice,  David,  1993:  Social  Skills  from  Mass  Higher  Education:  Rethinking  the  Company-Based 
Initial  Training  Paradigm.  Oxford  Review  of  Economic  Policy  9,  101-113. 

Soskice,  David,  forthcoming:  Divergent  Production  Regimes.  Coordinated  and  Uncoordinated  Market 
Economies  in  the  1980s  and  1990s.  To  appear  in  Herbert  Kitschelt,  Peter  Lange,  Gary  Marks 
and  John  D.  Stephens,  Continuity  and  Change  in  Contemporary  Capitalism.  Cambridge: 
Cambridge  University  Press. 

Sorensen,  Aage  B.,  and  Arne  L.  Kalleberg,  1981:  An  Outline  of  a Theory  of  the  Matching  of  Persons  to 
Jobs.  Pp.  49-74  in:  Ivar  Berg  (Ed.),  Sociological  Perspectives  on  Labor  Markets.  New  York: 
Academic  Press. 

Wegener,  Bernd,  1992:  Concepts  and  Measurement  of  Prestige.  Annual  Review  of  Sociology  18,  253- 
280. 


January  200 1 


Prepared  as  part  of  the  TSER  project: 

Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe 


Education  and  Labour  Market  Entry  across  Europe: 
The  Impact  of  Institutional  Arrangements 
In  Training  Systems  and  Labour  Markets 


Markus  Gangl 


Mannheim  Centre  for  European  Social  Research  (MZES) 

University  of  Mannheim 

P.O.Box  10  34  62 

D-68131  Mannheim 

Germany 

E-mail:  Markus.Gangl@mzes.uni-mannheim.de 


30G 

■ ‘V  ' ' 


WORKING  PAPERS 


Abstract 


Education  is  the  main  resource  of  young  people  entering  the  labour  market  for  securing  employment, 
in  competing  for  adequate  employment  contracts  and  in  fulfilling  occupational  aspirations.  As 
European  countries  differ  widely  in  the  institutional  structure  of  their  education  and  training  systems 
and  labour  markets,  different  resources  are  provided  to  school-leavers  entering  into  working  life  in 
different  countries,  who  additionally  face  varying  institutional  and  economic  contexts  in  labour  markets. 
The  paper  empirically  addresses  the  crucial  role  of  educational  qualifications  for  successful  labour 
market  entry  in  twelve  European  countries  in  the  mid-1990s,  drawing  on  the  1992-1997  European 
Community  Labour  Force  Survey.  The  main  aim  of  the  analyses  is  to  gauge  the  extent  to  which  cross- 
national differences  in  labour  market  outcomes  for  market  entrants  can  be  related  to  institutional 
differences  between  countries  in  terms  of  differences  in  qualification  profiles  of  school  leavers  and 
differences  in  terms  of  the  relationship  between  qualifications  and  early  labour  market  outcomes.  The 
analyses  cover  unemployment  and  occupational  allocation  as  two  major  dimensions  of  early  labour 
market  outcomes,  applying  multilevel  modelling  to  a database  of  repeated  comparative  cross-sectional 
surveys.  The  results  indicate  that  institutional  differences  in  both  education  and  training  systems  and 
labour  markets  play  a major  role  in  explaining  cross-national  differences  in  the  experiences  of  young 
people  entering  the  labour  market  in  EU  countries,  even  allowing  for  the  effects  of  variation  in 
economic  conditions  and  other  unmeasured  heterogeneity  between  countries  and  types  of 
qualifications. 


ERIC 


5.1 


1 Entering  the  Labour  Market  in  Europe: 

A Cross-National  Perspective 

It  is  by  now  widely  recognized  that  transition  processes  from  education  into  working  life  vary  markedly 
across  countries.  The  extent  to  which  young  people  entering  the  labour  market  are  subject  to  spells  of 
unemployment,  employment  in  specific  entry-level  occupations  and  industries,  or  prolonged  periods  of 
precarious  employment  situations  differs  markedly  among  European  Union  member  states  (OECD 
1996,  1998).  In  countries  like  France,  Greece,  Italy,  or  Spain,  youth  unemployment  is  a major  problem 
with  unemployment  rates  among  recent  school  leavers  amounting  to  more  than  30%  or  even  40%, 
sometimes  accompanied  by  massive  state  intervention  to  reduce  the  extent  of  the  problem.  Other 
countries  like  the  United  Kingdom  or  Ireland  see  less  of  a unemployment  problem,  but  there  are 
concerns  about  low  levels  of  training,  allocation  of  young  people  to  lower-level  jobs  and  excessive  job 
hopping  and  mobility  in  the  early  years  in  the  labour  force.  In  yet  another  set  of  countries,  notably 
Austria,  Denmark  or  Germany,  the  integration  of  young  people  into  the  labour  market  appears  much 
less  problematic.  There,  youth  unemployment  rates  are  generally  very  much  in  line  with  those  among 
more  experienced  workers  and  concerns  about  the  allocation  of  market  entrants  are  relatively  weak.  • 

There  is,  however,  abundant  evidence  that  these  cross-national  differences  in  labour  force  outcomes 
are  much  reduced  in  the  prime-age  labour  force  (Kerckhoff,  1995;  Layard  et  al.,  1991).  One  implication 
of  this  might  be  that  cross-national  differences  at  the  outset  of  employment  careers  reflect  the 
operation  of  distinct  institutional  arrangements  of  labour  market  entry  in  Europe,  providing  alternative 
mechanisms  for  integrating  young  people  into  the  labour  force  (Hannan  et  al.,  1999;  Muller  and  Shavit, 
1998;  Shavit  and  Muller,  2000,  forthcoming;  Kerckhoff,  1995,  forthcoming;  Allmendinger,  1989).  The 
institutional  structure  of  education  and  training  systems  has  long  since  been  considered  a likely 
explanation  for  cross-national  differences  at  labour  market  entry  in  general,  and  the  remarkably  better 
performance  of  countries  operating  (dual)  systems  of  occupationally-specific  training  at  the  secondary 
level  like  Austria,  Denmark,  the  Netherlands  or  Germany  in  terms  of  relatively  low  levels  of  both  youth 
unemployment  and  secondary  sector  employment  among  youth  (Hannan  et  al.,  1999;  Muller  and 
Shavit,  1998;  OECD  1998;  Allmendinger,  1989).  The  main  argument  in  this  body  of  literature  is  to 
relate  transition  patterns  and  outcomes  to  the  structured  integration  (Garonna  and  Ryan,  1989) 
occurring  through  the  provision  of  transferable  occupational  skills  and  extensive  work  experience  with 
a specific  employer  in  the  context  of  dual  training  system  arrangements,  mostly  apprenticeships, 
enabling  market  entrants  to  effectively  compete  for  jobs  in  occupationally  segmented  labour  markets 
(Marsden,  1986,  1990;  Marsden  and  Ryan,  1995).  In  countries  lacking  such  systems  of  training 
provision,  in  contrast,  early  labour  market  careers  are  said  to  exhibit  more  volatility,  unemployment 
and  job  mobility,  reflecting  more  extensive  periods  of  initial  job  search  and  the  acquisition  of  work 
experience  through  mobility  and  job  hopping  (Kerckhoff,  1993,  1995;  Rosenbaum  et  al.,  1990; 
Scherer,  1999). 

There  are  of  course  various  counter-arguments  challenging  the  validity  of  the  sketched  and  simplified 
institutional  account.  A first  one  might  consist  of  simply  acknowledging  the  more  favourable 
macroeconomic  conditions  in  the  set  of  dual  system  countries,  which  might  be  sufficient  to  explain  the 


better  prospects  of  market  entrants  there.  In  a similar  vein,  cross-national  differences  in  industrial  or 
occupational  structure  might  be  pointed  out  and  related  to  differences  in  industrial  or  occupational 
allocation  of  people  leaving  education  and  training.  Alternatively,  one  might  note  that  education  and 
training  systems  are  internally  heterogeneous  in  the  sense  that  the  characterization  of  a class  of  “dual 
system”-type  systems  conceals  that  any  such  differentiation  takes  the  part  of  upper  secondary  level 
training  to  represent  the  system  as  a whole  (Hannan  et  al.,  1999).  Hence,  even  if  correct,  the  above 
argument  might  in  fact  only  be  applicable  as  a partial  explanation  of  the  observed  cross-national 
differences,  as  long  as  additional  assumptions  relating  the  structure  of  training  systems  at  the  upper 
secondary  level  to  the  institutional  structure  of,  and  individual  behaviour  in,  other  parts  of  the  system 
are  not  explicitly  introduced.  Finally,  the  extent  to  which  the  above  argument  captures  the  empirically 
relevant  institutional  variation  among  European  Union  member  countries  might  itself  be  questioned,  as 
much  of  the  current  institutional  literature  is  developed  from  a decidedly  Northern  European 
perspective,  which  hardly  incorporates  Southern  European  patterns  into  the  systematic  argument. 

To  address  the  nature  and  sources  of  cross-national  differences  at  entering  the  labour  market  in  a 
more  precise  way,  the  paper  provides  a multilevel  analysis  of  unemployment  and  occupational 
allocation  among  market  entrants  in  twelve  European  countries,  drawing  on  the  1992-1997  European 
Community  Labour  Force  Survey.  At  the  individual  level,  the  main  focus  of  the  analyses  will  be  on  the 
role  of  education  for  successful  labour  market  entry  in  different  institutional  and  economic  context 
conditions,  while  a set  of  explicit  macrolevel  measures  of  the  latter  are  simultaneously  included  in  the 
analysis.  Hence,  the  relative  importance  of  several  alternative  explanations  for  the  observed  cross- 
national differences  can  be  assessed  empirically,  relating  different  patterns  of  labour  market  entry  to 
the  structure  of  qualificational  resources  of  market  entrants,  institutional  differences  in  the  association 
between  education  and  labour  market  attainment  or  varying  economic  context  conditions.  By  thus 
explicitly  addressing  the  micro-  and  macrolevel  aspects  of  institutional  effects  on  the  linkage  between 
education  and  initial  labour  force  outcomes,  the  paper  is  able  to  move  well  beyond  earlier,  purely 
macrolevel  accounts  (e.g.  OECD,  1998;  van  der  Velden/Wolbers,  2000;  but  cf.  also  Brauns  et  al., 
1998,  1999).  The  following  section  discusses  the  theoretical  background,  while  Section  3 describes 
the  database  and  research  design  applied  in  the  analyses.  The  empirical  results  are  discussed  in  two 
sections,  with  Section  4 providing  some  descriptive  information  and  Section  5 containing  the 
multivariate  analyses.  The  results  are  summarized  and  evaluated  in  the  concluding  section. 


ERIC 


5.3 


2 Education,  Labour  Market  Entry  and  the  Role  of  Context  Factors 


2.1  The  Role  of  Education  for  Successful  Entry  into  the  Labour  Market 

Education  is  the  main  resource  of  young  people  entering  the  labour  market  for  securing  employment, 
in  competing  for  adequate  employment  contracts  and  to  fulfill  their  occupational  aspirations  (Muller 
and  Shavit,  1998;  Hannan  et  al.,  1999).  Education  provides  both  productive  capacities  to  individuals 
and  signals  of  these  to  potential  employers  (Breen  et  al.,  1995;  Becker,  1993;  Spence,  1973,  1981; 
Bills,  1988;  Hunter  and  McKenzie  Leiper,  1993;  Spilerman  and  Lunde,  1991;  Polacheck  and  Siebert, 
1993;  Ashton  and  Sung,  1992)  - hence,  attained  qualifications  are  a main  asset  in  worker  competition 
for  jobs  available  in  the  labour  market.  Of  course,  education  is  not  the  only  resource  of  workers  in  job 
search:  work  experience,  past  employment  history,  networks  and  contacts,  or  geographical  mobility 
might  all  be  reasonably  and  convincingly  related  to  individuals’  labour  market  success.  In  addition, 
social  differentiation  according  to  gender,  ethnicity  or  class  background  might  be  expected  to  operate, 
both  due  to  their  association  with  the  availability  of  market  resources  and  more  fundamental  persisting 
social  inequalities.  There  are,  however,  at  least  two  reasons  which  encourage  a systematic  focus  on 
the  role  of  education  in  analysing  the  transition  from  education  to  work.  First  of  all,  most  of  these 
factors  like  work  experience,  employment  history  or  professional  contacts  are  zero  or  at  least  very 
limited  among  those  entering  the  labour  force  almost  by  definition.  That  is,  to  the  extent  that  labour 
market  allocation  depends  on  factors  other  than  educational  qualifications,  labour  market  entrants  are 
among  the  least  competitive  job  seekers  as  they  necessarily  lack  these.  In  contrast,  young  people 
entering  the  labour  market  have  invested  in  their  qualifications,  at  least  in  part  in  order  to  achieve 
adequate  employment  prospects  at  later  life  stages.  At  entering  the  market,  this  training  process  is 
completed  (if  only  temporarily)  as  sufficiently  satisfactory  qualifications  have  been  obtained  and 
individuals  aim  to  extract  labour  market  returns  to  these.  To  the  extent  that  labour  market  processes 
depend  on  the  qualification  attained,  early  labour  force  experiences  are  an  immediate  consequence  of 
educational  decisions  taken  earlier,  and  thus  intimately  linked  together  and  potentially  both  mutually 
reinforcing  and  behaviourally  interdependent.  To  sum  up  the  arguments,  assessing  early  labour  force 
outcomes  provides  an  estimate  of  short-term  returns  (in  biographical  terms)  to  educational 
investments  which  young  people  can  expect  on  the  labour  market. 


Such  labour  market  returns  to  educational  investments  are  conceptually  most  easily  understood  in  the 
framework  of  hiring  and  market  allocation  models  in  the  spirit  of  Thurow  (1975)  or  Rosen  (1972;  cf. 
also  the  related  sociological  literature  on  labour  market  matching  following  Sorensen  and  Kalleberg, 
1981).  To  simplify,  it  is  assumed  in  these  and  related  models  that  employers’  readiness  to  hire  an 
individual  into  a specific  employment  position  depends  on  the  expected  training  costs  of  the  individual 
should  it  be  employed  at  that  position.  That  is,  the  smaller  the  differential  between  expected  current 
individual  productivity  and  required  productivity  in  the  position  in  question,  the  larger  the  likelihood  of  a 
hire.  A number  of  aspects  of  educational  qualifications  are  obviously  related  to  and  provide  signals 
about  different  components  of  productivity  and,  in  turn,  expected  training  costs.  Regularly,  the  current 
literature  identifies  the  level  and  the  vocational  specificity  (or  occupational  specialization)  of 


O 

ERIC 


5.4 


3in 


qualifications  as  two  main  dimensions  of  relevance  (cf.  the  contributions  in  Shavit  and  Muller,  1998; 
Braun  and  Muller,  1998;  Brauns  and  Steinmann,  1999).  Both  the  level  of  qualifications  as  an  index  of 
general  ability  and  the  vocational  specialization  of  training  received  as  an  indicator  of  transferable 
occupationally-specific  skills  can  be  expected  to  reduce  incurred  training  costs,  and  thus,  to  increase 
the  probability  of  access  to  employment,  with  the  effect  being  mostly  confined  to  the  sector  of 
specialization  in  the  latter  case.  But  the  discussion  of  different  types  of  national  institutional  systems  of 
training  provision  suggests  this  list  is  not  of  institutional  variation  relevant  to  cross-national  research 
(cf.  the  review  in  Hannan  et  al.,  1999).  More  specifically'  in  those  European  countries  regularly 
credited  the  most  successful  ones  in  terms  of  youth  integration  into  the  labour  force,  it  is  fairly  common 
to  provide  vocational  training  in  an  environment  combining  school-  and  work-based  training,  notably  in 
the  form  of  apprenticeship  systems.  This  type  of  training  adds  a third  dimension  of  importance  to  the 
picture  which  might  be  phrased  as  provision  of  work  experience  and  direct  employer  involvement  in 
training  provision  (Hannan  et  al.,  1999).  With  respect  to  expected  training  costs,  such  training  provides 
at  least  two  additional  advantages  to  market  entrants:  work  experience  and  firm-specific  training, 
which  can  both  be  expected  to  reduce  the  expected  productivity  differential,  though  the  latter  should 
apply  to  continued  employment  with  the  training  firm  only.  In  sum,  these  arguments  amount  to  the 
following  three  hypotheses  on  the  relationship  between  educational  qualifications  and  unemployment 
risks  of  market  entrants,  which  is  for  simplicity  conceptualized  as  the  probability  of  non-hiring:  (HYP1) 
Unemployment  is  negatively  associated  with  the  level  of  education:  (HYP2)  at  a given  level  of 
education,  unemployment  risks  are  lowered  by  attaining  qualifications  which  provide  vocational 
specialization:  and  (HYP3)  at  a given  level  of  education,  unemployment  among  market  entrants  is 
lowered  by  completing  an  apprenticeship  or  training  in  a similar  type  of  dual  system  arrangement. 

The  reasoning  easily  extends  to  expectations  about  the  relationship  between  qualifications  and 
occupational  attainment.  Following  Thurow’s  model  again,  one  might  imagine  a ranking  of  available 
jobs  according  to  general  attractivity,  with  remuneration  and  required  productivity  being  relatively 
closely  linked.  If  individuals  strive  to  attain  highly  rewarded  employment  positions,  they  will  accept 
employment  at  the  most  attractive  employment  position  available.  As  the  availability  of  positions  is 
assumed  to  depend  on  expected  training  costs,  the  choice  set  is  increasingly  restricted  at  lower 
qualification  levels.  From  this  it  follows  that  (HYP4)  with  respect  to  occupational  allocation  along  the 
reward  hierarchy,  it  should  only  be  the  level  of  education  which  is  of  importance,  rather  than  vocational 
specificity  of  qualifications  in  itself.  The  vocational  specialization  of  qualifications  would,  in  contrast,  be 
expected  to  affect  the  probability  of  employment  within  the  sector  of  specialization,  as  training  costs 
are  by  definition  only  reduced  in  that  specific  sector.  Given  data  limitations  with  respect  to  the  level  of 
detail  in  the  measurement  of  qualifications,  however,  it  is  not  possible  to  pursue  this  last  issue  in  the 
empirical  analyses  to  follow. 


ERIC 


5.5 


2.2  The  Importance  of  Institutional  Contexts 

While  the  above  reasoning  attempted  to  provide  a brief  outline  of  the  general  mechanisms  linking 
education  and  market  entrants’  labour  force  outcomes,  a main  strength  of  comparative  research  is  the 
ability  to  address  the  impact  of  fundamentally  varying  context  conditions  between  countries  on  that 
nexus.  Of  primary  relevance  are  the  institutional  structure  of  education  and  training  systems  on  the 
one  hand  and  labour  markets  on  the  other,  where  crucial  parameters  affecting  individual  decisions  and 
their  aggregate  outcomes  in  the  transition  process  from  education  to  work  are  set.  European 
countries,  which  are  considered  here,  differ  markedly  in  both  respects,  so  that  wide  cross-national 
differences  in  transition  outcomes  might  be  expected. 

The  institutional  structure  of  education  and  training  systems  is  at  the  center  of  many  explanatory 
frameworks  addressing  cross-national  variation  in  the  nature  of  early  labour  market  careers,  although 
specific  arguments  vary  widely  (cf.  Hannan  et  al.,  1999;  Muller  and  Shavit,  1998,  2000;  Allmendinger, 
1989;  Kerckhoff,  1995,  forthcoming).  It  is  certainly  true  that  institutional  arrangements  in  education 
systems  are  of  fundamental  importance  to  transition  outcomes  as  they  channel,  constrain  or  enable 
sufficient  individual  acquisition  of  qualifications.  If  one  views  the  nature  of  qualifications  individuals 
have  achieved  by  entering  the  labour  market  as  the  outcome  of  a basically  rational  decision  on  the 
part  of  young  people  and  their  families  (Becker,  1993;  Breen  and  Goldthorpe,  1997;  Jonsson,  1999; 
Borghans  and  Groot,  1999),  then  it  is  self-evident  that  the  institutional  structure  of  training  systems  can 
be  seen  as  defining  the  educational  choice  set  and  the  properties  of  the  discrete  qualificational 
alternatives  provided,  which  might  be  assessed  in  terms  of  expected  costs  and  benefits.  Phrased  in 
this  simple  framework,  institutional  variation  in  education  and  training  systems  might  consist  of  both 
differences  in  the  choice  sets  offered  and  the  features  of  specific  qualifications.  To  the  extent  that  such 
variation  occurs,  one  would  expect  to  observe  variation  in  the  educational  distribution  of  market 
entrants,  which  is  indeed  widely  documented  for  European  countries  (e.g.  Muller  and  Wolbers,  1999). 
A main  result  of  these  studies  is  to  point  out  wide  cross-national  variation  in  the  availability  of 
apprenticeships  or  other  dual  systems  of  training.  In  countries  like  Germany,  Austria  or  Denmark  such 
training  appears  to  provide  an  attractive  initial  qualification  to  young  people,  while  such  training 
environments  have  declined  - for  numerous  reasons  (e.g.  Marsden  and  Ryan,  1995)  - in  importance  in 
the  Netherlands  or,  much  earlier,  Britain,  and  are  only  recently  returning  to  some  degree  in  France  or 
through  the  Modern  Apprenticeship  Programme  in  the  United  Kingdom.  On  the  other  hand,  European 
countries  differ  markedly  in  the  extent  to  which  youth  cohorts  attain  upper  secondary  or  tertiary  level 
qualifications,  with  e.g.  Austria  and  Italy  exhibiting  very  low  and  hardly  increasing  proportions  of 
tertiary  level  leavers  or  the  proportions  of  young  people  leaving  the  educational  system  with  essentially 
compulsory  schooling  continuing  to  be  relatively  high  in  Southern  Europe,  but  also  in  Britain  (OECD 
1997;  Muller  and  Wolbers,  1999;  Muller  et  al.,  1997).  Given  that  different  educational  systems  thus 
provide  market  entrants  with  very  different  sets  of  qualificational  resources,  cross-national  differences 
in  initial  labour  market  outcomes  and  the  nature  of  the  transition  process  itself  are  to  be  expected. 
More  specifically,  the  total  effects  of  education  and  training  systems  on  the  transition  from  school-to- 
work  are  restricted  to  compositional  differences  in  market  entrants’  stock  of  qualifications  at  leaving 
the  training  system.  Consequently,  it  is  hypothesized  that  (HYP5)  cross-national  differences  in  early 


ERIC 


labour  market  outcomes  arise  from  cross-national  variation  in  training  systems'  effectiveness  in 
providing  young  people  with  qualificational  resources  valued  in  the  labour  market.  Those  countries 
especially  successful  in  avoiding  high  rates  of  youth  unemployment  and  high  rates  of  inadequate 
employment  among  market  entrants  to  a larger  extent  enable  young  people  to  acquire  the  necessary 
qualifications  for  such  sucessful  market  entry.  That  is,  conforming  to  the  microlevel  hypotheses  set  out 
above,  it  is  expected  that  those  educational  systems  providing  advantageous  - i.e.,  tertiary  level, 
vocationally  specific  or  apprenticeship  - qualifications  to  a larger  share  of  a cohort  of  market  entrants 
lead  to  more  favourable  patterns  of  youth  integration  into  the  market. 

European  countries  also  differ,  however,  in  the  institutional  structure  of  labour  markets  young  people 
have  to  face  at  market  entry,  and  this  might  be  equally  expected  to  impact  on  the  nature  of  education- 
to-work  transitions  in  the  different  countries.  Of  course,  there  are  a multitude  of  angles  from  which  the 
issue  of  labour  market  institutions  might  be  addressed,  focusing  e.g.  on  the  nature  of  formal 
employment  regulation,  union  bargaining  power  and  the  nature  of  wage  bargaining  systems,  the 
nature  of  labour  market  segmentation  or  the  extent  of  active  labour  market  programmes  aimed  at 
integrating  young  people  into  the  work  force,  all  of  which  may  easily  be  related  to  expectations  about 
transition  outcomes  (e.g.  van  der  Velden  and  Wolbers,  2000b).  Attempting  to  identify  the  main 
institutional  features  of  relevance  from  a small  sample  of  countries  is,  however,  plagued  by  the  fact  of 
institutional  interdependencies  in  the  above  and  other  characteristics  of  labour  markets,  implying  both 
a theoretical  indeterminacy  of  results  and  patterns  of  severe  multicollinearity  in  the  empirical  data.  As 
the  focus  of  the  paper  is  on  the  role  of  education  and  educational  systems,  the  potential  additional 
explanatory  power  of  institutional  arrangements  in  labour  markets  is  assessed  in  more  explorative 
ways  in  the  following.  Building  on  earlier  empirical  analyses  (Gangl,  2000a),  three  country  clusters  are 
distinguished  in  the  set  of  twelve  EU  member  countries,  representing  both  distinct  configurations  of 
institutional  arrangements  and  empirical  patterns  of  labour  market  entry.  More  specifically,  I distinguish 
(a)  Austria,  Denmark,  Germany  and  the  Netherlands  as  the  group  of  Northern  European  countries 
where  strong  occupational  labour  markets  are  regularly  claimed  to  operate  (Muller  and  Shavit,  1998; 
Eyraud  et  al.,  1990;  Marsden,  1990),  (b)  the  Southern  European  countries,  including  Greece,  Italy, 
and  Portugal;  regularly  claimed  to  exhibit  rigid  labour  market  context,  both  in  terms  of  formal 
employment  protection  and  career  mobility  patterns  (Grubb  and  Wells,  1993;  OECD  1999;  Bernardi  et 
al.,  1999;  Jobert,  1997),  while  treating  (c)  the  residual  set  of  countries,  including  Britain,  Ireland, 
France,  Belgium,  but  also  Spain,  as  a final  group  of  European  countries,  which  (to  varying  degrees) 
lack  either  institutional  feature  of  labour  markets.  Following  standard  practice  in  the  literature,  the 
labels  of  OLM  systems,  Southern  European  systems,  and  ILM  systems  will  be  used  as  shorthands  for 
these  groups,  respectively.1  In  more  theoretical  terms,  this  distinction  alludes  to  different  institutional 
solutions  (even  if  implicit)  of  reducing  the  productivity  differential  for  market  entrants,  which  may 


1 I explicitly  note  the  potentially  misleading  use  of  these  labels  as  a singular  feature  is  used  to  characterize 
types  of  systems.  The  intended  use  here  is  rather  one  of  relatively  similar  institutional  arrangements  shaping 
transition  processes,  consisting  of  fairly  distinct  and  internally  relatively  coherent  combinations  of  institutional 
features  in  labour  markets.  Such  institutional  interdependencies  might  also  include  relations  between  the 
structure  of  training  systems  and  the  structure  of  labour  markets  (e.g.  Maurice  and  Sellier,  1986;  Marsden, 


ERIC 


5.7 


313 


heuristically  be  contrasted  along  the  two  axes  of  skill  vs.  wage/contract  flexibility  in  labour  markets. 
One  strategy  to  increase  young  people’s  competitiveness  on  the  labour  market  is  early  skill 
specialization  aimed  at  improving  initial  relative  productivity  of  market  entrants  as  compared  to  that  of 
adult  workers.  This  strategy  is  chosen  in  the  (ideal-typical)  occupationalized  systems  of  the  first  group 
of  countries,  which  combine  strong  vocationally-oriented  training  systems  with  both  fairly  regulated 
and  occupational  labour  markets  (cf.  the  lower  right  hand  panel  in  Table  1).  Alternatively,  youth  labour 
market  integration  might  be  achieved  by  flexibilizing  labour  usage  and  employment  contracts  in  order 
to  allow  for  a closer  relation  between  market  entrants’  current  productivity  and  job  rewards  in  terms  of 
pay  and  other  contract  conditions.  This  ideal-typical  second  strategy  is  operated  in  the  other  North 
European  systems  in  the  upper  left  hand  cell  of  Table  1,  where  labour  market  regulation  is  weak  in 
general  (as  in  the  UK  or  Ireland)  or  deliberately  flexibilized  by  introducing  wage  subsidies,  work 
experience  programmes  and  flexible  forms  of  contracts  among  young  people  (as  in  France,  Belgium 
or  Spain).  A second  component  of  youth  labour  market  integration  in  this  context  might  consist  of 
disproportionally  allocating  young  people  into  the  secondary  sector  work  force  and  subsequent 
promotion  from  this  pool  of  workers  afterwards  (Marsden,  1990).  In  some  sense,  either  attempt  of 
fostering  labour  market  integration  of  young  people  appears  to  be  absent  in  the  remaining  set  of 
Southern  European  countries  assembled  in  the  lower  left  hand  cell  of  Table  1.  In  consequence,  one 
might  expect  that  (HYP6)  unemployment  risks  among  market  entrants  should  be  markedly  more 
pronounced  in  the  latter  systems  as  compared  to  both  OLM  and  ILM  systems,  while  (HYP7)  the  main 
contrast  between  OLM  and  Southern  systems  as  compared  to  ILM  systems  should  be  a more 
favourable  occupational  allocation  at  market  entry. 


Table  1 Types  of  Institutional  Arrangements  in  Labour  Markets 


Wage/Contract 

Flexibility 

Skill  Flexibility 

high 

low 

high 

United  Kingdom,  Ireland 

Spain,  France,  Belgium 

the  Netherlands,  Denmark 

low 

Greece,  Italy,  Portugal 

Austria,  Germany 

Source:  based  on  empirical  results  in  Gang I (2000a) 


1990;  Muller  and  Shavit,  1998;  Hannan  et  al.,  1999),  which  render  the  analytical  separation  between  the  two 
somewhat  less  unequivocal  than  discussed  above. 

5.8 


2.3  Economic  Context  and  Labour  Market  Entry 


Labour  markets  are  not  continuously  in  stable  equilibrium,  but  rather  constantly  adapting  and  adjusting 
to  various  sources  of  change.  These  are  both  short-term  forces  like  labour  market  and  employment 
reactions  to  business-cycle  fluctuations  in  product  markets  (e.g.  Blossfeld,  1986;  Storer,  1994;  Bowlus, 
1995),  but  also  more  long-term  changes  in  the  skill  structure  of  employment  or  the  qualificational 
structure  of  labour  supply  (Muller  and  Wolbers,  1999;  Brauns,  1998;  Muller  et  al.,  1997;  Dronkers, 
1999;  Penn  et  al.,  1994;  Gregg  and  Manning,  1997;  Evans,  1999;  Ashton  et  al.,  1990),  which  have  the 
labour  market  must  become  accustomed  by  to,  and  change  accordingly,  over  time.  Hence,  cross- 
national variation  in  labour  market  outcomes  always  reflects  effects  of  institutional  differences  between 
countries  as  well  as  country  differences  with  respect  to  general  economic  conditions  or  other  national 
labour  market  factors  like  a particular  industrial  or  occupational  structure.  In  order  to  properly  identify 
genuine  institutional  effects  in  empirical  analysis,  one  has  to  allow  for  any  such  effects  of  the  state  of, 
and  changing  economic  contexts  in,  the  labour  market.  It  is  easily  imagined  that  general  economic 
conditions  affect  market  entrants’  labour  force  outcomes,  so  that  cross-national  variation  in  transition 
outcomes  reflect  varying  aggregate  labour  market  conditions.  I will  not  proceed  to  develop  any 
stringent  tests  of  economic  context  effects  as  they  are  addressed  more  specifically  in  a companion 
paper  to  the  current  one  (Gangl,  2000b).  Rather  it  should  be  sufficient  to  note  some  general 
expectations  about  these  effects  without  presenting  more  elaborate  theoretical  justifications  at  this 
point.  In  general,  the  hypothesis  is  that  young  people’s  unemployment  risks  will  be  negatively  related 
to  aggregate  labour  market  conditions  as  measured  by  unemployment  rates  in  the  total  labour  force 
and  positively  to  increasing  youth  cohort  sizes  as  captured  by  the  youth-adult-ratio  in  the  labour  force. 
Occupational  allocation,  in  contrast,  is  expected  to  depend  mainly  on  changes  in  the  skill  balance  in 
the  market,  with  educational  expansion  leading  to  decreasing  occupational  outcomes  while  increasing 
professionalization  of  the  work  force  is  supposed  to  imply  rising  levels  of  occupational  allocation.  In  the 
current  context,  it  is,  however,  mainly  relevant  that  the  database  used  in  the  empirical  analyses  allows 
usto  introduce  such  variables  as  additional  controls,  potentially  enabling  us  to  arrive  at  clearer 
estimates  of  genuine  institutional  effects  on  transition  outcomes  in  various  European  countries.  I now 
turn  to  describe  the  data  in  more  detail. 


Following  the  above  introduction,  this  paper  will  present  a set  of  comparative  analyses  of  labour 
market  entry  in  the  countries  of  the  European  Union  in  the  mid-1990s.  In  the  analyses,  data  for  twelve 
European  countries  is  used,  drawing  on  the  1992-1997  European  Community  Labour  Force  Surveys.2 


2 This  data  has  kindly  been  provided  by  EUROSTAT,  the  Statistical  Office  of  the  European  Union.  Of  course, 
EUROSTAT  is  neither  responsible  for  the  uses  made  of  the  data  nor  the  views  held  by  the  author.  The  twelve 
countries  chosen  for  analyses  are  Austria,  Belgium,  Denmark,  France,  Germany,  Greece,  Ireland,  Italy,  the 
Netherlands,  Portugal,  Spain  and  the  United  Kingdom.  Luxembourg  is  excluded  for  reasons  of  small  sample 
sizes  giving  unreliable  results,  while  Finland  and  Sweden  had  to  be  excluded  as  occupational  information  is 
only  provided  in  1997.  For  the  chosen  set  of  countries,  single  annual  observations  were  excluded  due  to 
. breaks  in  (part  of)  the  time  series  or  other  unreliabilities,  mostly  related  to  substantial  changes  in  the  coverage 


3 Data  and  Methodology 


ERIC 


5.9 


This  database  provides,  standardised,  cross-sectional  information  on  labour  force  participation, 
unemployment  and  various  aspects  of  employment  compiled  from  EU  member  states'  national  Labour 
Force  Surveys.3  The  surveys  themselves  consist  of  large-scale  national  samples  which  are  repeated 
at  least  annually,  thus  providing  a unique  database  of  repeated  cross-sectional  surveys  of  labour 
market  behaviour  and  employment  issues  in  EU  countries  (cf.  EUROSTAT  1992,  1996,  for  extensive 
details  on  the  database). 

For  the  analyses,  a subsample  of  market  entrants  is  drawn  from  the  full  ECLFS  database:  market 
entrants  are  defined  as  those  individuals  as  who  left  formal  education  and  training  no  longer  than  five 
years  ago.  As  individual  time  of  leaving  education  is  unavailable  in  the  database,  the  timing  of  market 
entry  is  proxied  by  typical  graduation  ages  for  the  different  levels  and  types  of  education  as  published 
by  the  OECD  (1997).  At  the  individual  level,  gender,  potential  labour  force  experience  and  level  and 
type  of  education  is  observed.  Potential  labour  force  experience  is  measured  in  years  since  the 
OECD’s  age  of  typical  graduation  for  the  highest  level  and  type  of  education  achieved.  Highest  level 
and  type  of  education  achieved  is  measured  in  terms  of  an  augmented  ISCED  classification  (UNESCO 
1975)  which  distinguishes  four  levels  of  qualifications  as  present  in  the  standard  ISCED  scheme,  but 
supplementing  this  by  differentiating  three  types  of  qualifications  at  the  upper  secondary  level.  More 
specifically,  the  qualification  levels  distinguished  are:  ISCED  levels  0-2  or  having  attained  no  more 
than  lower  secondary  qualifications,  ISCED  level  3 or  having  attained  upper  secondary  education, 
ISCED  level  5 or  having  attained  post-secondary  or  lower  tertiary  qualifications,  and  ISCED  levels  6-7 
or  having  attained  full  university  or  Ph.D.  degrees.  In  addition,  the  level  of  upper  secondary  education 
(ISCED  3)  is  further  subdivided  according  to  the  nature  of  qualifications  into  upper  secondary  general 
academic  education,  upper  secondary  school-based  vocational  training,  and  apprenticeship  training, 
yielding  six  educational  groups  in  total.4  All  variables  are  measured  as  of  an  individually-specific 
reference  week,  regularly  scheduled  in  spring  each  year. 

Based  on  the  ECLFS  dataset,  unemployment  risks  and  occupational  attainment  are  analyzed  as  two 
main  aspects  of  early  labour  market  attainment  (unfortunately,  neither  earnings,  wages  nor  income  is 
available  from  the  ECLFS).  With  respect  to  employment  and  unemployment,  the  ECLFS  database 
follows  standard  international  ILO  definitions  (cf.  ILO,  1990a),  while  occupations  are  classified 
according  to  the  ISCO-88  COM  scheme  at  the  3-digit  level  (cf.  ILO  1990b,  EUROSTAT  1992,  1996). 
In  the  current  paper,  a small  modification  to  the  ILO  concept  of  employment  is  applied:  in  an  analysis 
of  early  labour  market  experiences  and  the  transition  from  education  to  work,  it  appears  unwarranted 
to  include  all  individuals  having  worked  for  payment  or  profit  without  paying  attention  to  any  current 
participation  in  education  and  training,  which  might  actually  represent  their  primary  status.  Deviating 
from  standard  ILO  procedures,  all  individuals  participating  in  any  kind  of  initial  formal  education  and 


of  current  training  activities  or  the  coding  of  educational  qualifications  (e.g.  France,  Belgium  or  Ireland  1992). 
Data  on  Austria  is  only  available  since  1995,  when  that  country  joined  the  European  Union. 

Standardisation  of  information  closely  adheres  to  established  international  standards  as  laid  down  in  ILO 
conventions  (cf.  ILO  1990;  EUROSTAT  1992,  1996). 

Additional  individual-level  information  is  present  in  the  database,  yet  unavailable  for  scientific  research  as 
current  data  protection  policies  restrict  access  to  nine-dimensional  multivariate  tables. 


316 


5.10 


training  are  therefore  excluded  from  the  labour  force.5  After  all,  market  entrants  are  thus  defined  as 
individuals  having  (intermittently,  if  only)  completed  their  educational  careers.  For  this  group,  labour 
market  outcomes  are  investigated  in  terms  of  unemployment  risks  in  early  careers  and  initial 
occupational  allocation,  the  latter  being  measured  in  terms  of  occupational  status,  the  incidence  of 
lower-skilled  employment  and  the  attainment  of  professional  employment  positions.  Below,  Overview  1 
provides  more  specific  details  on  the  measurement  of  each  concept. 


Overview  1 

Employment 
& Labour  Force 

Modified  ILO  international  definition  of  employment  / labour  market 
participation  (cf.  ILO  1990a):  participation  in  initial  training  considered 
as  primary  status 

Unemployment 

ILO  international  definition  of  unemployment  (cf.  ILO  1990a) 

Occupational  Status 

ISEI  international  socio-economic  index  score  (cf.  Ganzeboom  et  al., 
1992;  1996)  matched  at  the  level  of  3-digit  ISC088-C0M  occupational 
detail 

Lower-skilled  Employment 

Un-/semi-skilled  or  lower-level  occupation  according  to  ISCO 
classification  (cf.  ILO  1990b):  ISC088-C0M,  3-digit  occupational 
groups  422,  512,  520,  522,  611-615,  822-830,  832-933  (e.g.  lower- 
level  salesworkers,  restaurant  workers,  machine  operators,  drivers, 
elementary  services  occupations,  agricultural  and  production 
labourers) 

Professional  Employment 

Professional  occupation  according  to  ISCO  classification  (cf.  ILO 
1990b):  ISC088-C0M,  2-digit  occupational  groups  11,  12,  21-33  (e.g. 
teaching  and  scientific  professionals,  managers,  architects,  health 
professionals,  technicians) 

In  the  current  analyses,  these  individual-level  measures  are  complemented  by  a set  of  context  factor 
measures,  which  are  conceived  of  as  including  both  institutional  variables  and  labour  market  context 
factors.  As  discussed  earlier,  the  latter  set  of  variables  is  basically  introduced  as  additional  contextual 
controls  which  will  not  be  discussed  explicitly  in  what  follows  below  (but  see  Gangl,  2000b,  which  even 
uses  a full  decade  of  information  from  the  same  database).  The  four  macrolevel  measures  included  in 
the  analyses  are  (a)  the  demographic  size  of  youth  cohorts  in  terms  of  the  youth-adult  ratio  in  the 
labour  force,  i.e.  the  ratio  of  market  entrants  to  experienced  workers  in  the  total  labour  force  aged  15- 
59,  following  the  sample  specifications  detailed  above;  (b)  the  aggregate  unemployment  rate  in  the 
total  labour  force  aged  15-59,  indexing  aggregate  economic  conditions  and  business  cycle 
fluctuations,  (c)  the  extent  of  educational  expansion  as  captured  by  the  proportion  of  tertiary-level  - i.e. 
ISCED  5-7  - qualified  individuals  in  the  total  labour  force,  and  (d)  the  extent  of  labour  market 
professionalization  as  measured  by  the  proportion  of  professional  employment  positions  (as  defined  in 


This  rather  open  and  imprecise  formulation  is  meant  to  include  participation  in  those  regulated  forms  of 
training  which  might  be  considered  in  some  way  as  ‘initial’,  while  excluding  those  types  of  of  part-time 
education  which  serve  to  enhance  individual  qualifications  while  already  working.  Attending  university,  upper 
secondary  schools  or  dual-system  types  of  training  would  be  examples  of  the  former;  attending  evening 
schools  or  firm-based  training  courses  examples  for  the  latter.  Full  details  of  coding  are  available  from  the 
author  on  request. 


5.11 


317 


Overview  1)  among  total  employment.  All  measures  are  based  on  estimates  from  the  ECLFS 
database  for  66  country  level  observations,  i.e.  12  sample  countries  times  3-6  annual  observations. 
The  analyses  themselves  utilize  both  within-country  mean-centered  values  to  characterize  within- 
country  changes  in  economic  contexts,  and  centered  mean  values  to  capture  stable  between-country 
differences  in  the  period  of  observation.  As  introduced  in  the  theoretical  section  above,  the  variable 
characterizing  institutional  labour  market  contexts  amounts  to  a simple  differentiation  between  the 
three  country  clusters  discussed  earlier.  The  impact  of  education  and  training  institutions  on  transition 
outcomes  will  implicitly  be  controlled  in  the  statistical  analyses  by  conditioning  labour  market  outcomes 
on  types  of  education  attained. 

In  the  following,  the  paper  will  first  present  descriptive  evidence  on  both  cross-national  variation  in 
early  labour  market  outcomes  and  the  role  of  educational  qualifications  for  unemployment  risks  at 
entering  the  labour  market  and  initial  occupational  allocation  in  the  twelve  EU  countries  under  study. 
Early  labour  market  outcomes  will  then  be  assessed  from  comparative  micro-macro  models, 
controlling  simultaneously  for  individual  resources,  institutional  and  economic  context  factors,  and 
unobserved  heterogeneity  between  countries  and  qualifications.  This  modelling  strategy  follows  in  a 
straightforward  manner  from  the  repeated  cross-sectional  research  design  of  the  database  used  (cf. 
Blalock,  1984;  Judge  etal.,  1985;  Greene,  1997;  DiPrete/Grusky  1990a,  1990b)  and  applies  multilevel 
or  generalized  linear  mixed  model  estimation  in  the  analyses  (cf.  Bryk  and  Raudenbush,  1992; 
Longford  1993,  1995;  Goldstein,  1995).  All  models  control  for  fixed  effects  of  the  set  of  covariates 
introduced  above,  but  include  the  estimation  of  two  normally  distributed  random  effects,  one  for  the 
country  level  and  a second  for  the  more  than  60  national  qualifications  distinguished  in  order  to 
account  for  unmeasured  heterogeneity  between  countries  and  types  of  qualifications  (cf.  van  der 
Velden  and  Wolbers,  2000,  for  a similar  application).  The  calculation  of  standard  errors  and  hypothesis 
tests  is  adjusted  for  the  clustering  of  observations  by  country  and  type  of  education  within  country.  The 
dichotomous  dependent  variables  of  unemployment,  lower-skilled  employment  and  professional 
employment  are  modelled  by  specifying  a binomial  distribution  and  a logit  link  function,  occupational 
status  is  specified  to  follow  a normal  distribution  with  a logarithmic  link  function. 


4 Education  and  Early  Labour  Market  Outcomes  across  Europe 

As  a first  step  in  the  empirical  analyses,  some  descriptive  evidence  on  the  relationship  between 
market  entrants’  educational  background  and  their  initial  labour  market  outcomes  will  be  briefly 
presented.  The  following  Figures  1-4  provide  simple  cross-tabulations  between  education  and  the  four 
outcome  indicators  on  unemployment  and  occupational  allocation  for  each  of  the  twelve  countries  in 
the  sample,  averaged  across  the  period  1992-1997.  Evidently,  it  is  impossible  to  discuss  all  aspects  of 
these  results  in  detail,  so  an  attempt  will  be  made  to  emphasize  some  broad  tendencies,  while  the 
multivariate  analyses  in  the  following  section  aim  to  provide  a much  more  condensed  account  of  the 
systematic  components  shaping  labour  market  entry  outcomes  in  Europe. 


Turning  to  market  entrants’  unemployment  risks  first,  the  results  clearly  show  substantial  variation, 
both  between  countries  and  between  types  of  education,  but  also  between  equivalent  types  of 
education  across  countries.  Unemployment  risks  are  lowest,  in  general,  in  those  four  countries 
operating  large  scale  occupationally-specific  training  systems:  unemployment  rates  range  from  7%  in 
Austria  to  about  10%  in  the  Netherlands,  Denmark  and  Germany.  In  Portugal,  Belgium,  the  United 
Kingdom  and  Ireland,  unemployment  rates  are  between  13%  and  21%,  while  they  amount  to  about 
30%  in  France  and  Greece,  37%  in  Spain,  and  even  42%  in  Italy.  Of  course,  there  is  equally  wide 
variation  between  types  of  education:  university  leavers  in  Austria  face  an  unemployment  risk  of  5% 
only,  while  half  of  all  entrants  from  upper  secondary  general  or  lower  tertiary  education  in  Italy  or  lower 
secondary  education  in  France  are  unemployed.  In  general,  unemployment  rates  decline  with 
increasing  levels  of  qualifications.  In  France,  for  example,  unemployment  rates  at  the  upper  secondary 
level  are  about  half  the  figure  for  the  lowest  qualified,  and  even  reduced  to  about  one  third  for  leavers 
from  tertiary  level  education.  Similar  relations  hold  in  countries  like  the  United  Kingdom,  Ireland, 
Denmark  and  the  Netherlands.  Moreover,  apprenticeships  evidently  perform  very  favourably,  both 
compared  to  school-based  education  at  the  same  level  of  training  and  across  qualificational  levels.  In 
all  countries  operating  any  such  training,  unemployment  rates  for  apprentices  tend  towards  those  of 
tertiary  level  leavers  and  are  certainly  far  from  those  of  leavers  with  compulsory  education  only.  But 
there  is  additional  variation  between  the  same  type  of  education  across  countries:  notably,  Southern 
European  countries  are  distinct  in  the  sense  that  there  are  hardly  any  benefits  attached  to  achieving 
higher  qualification  levels  in  terms  of  unemployment.  Rather,  unemployment  rates  in  Southern  Europe 
are  often  even  higher  at  the  upper  secondary  level  as  compared  to  those  of  the  lowest  qualified 
leavers,  and  not  much  reduced  for  tertiary  level  graduates  either. 

Occupational  allocation,  in  turn,  is  addressed  by  the  three  indicators  of  occupational  status  attainment, 
the  probability  of  lower-skilled  employment  and  the  probability  of  employment  in  the  professions  and 
semi-professions,  respectively.  In  terms  of  occupational  status  attainment  there  is  little  evidence  of 
cross-national  variation,  neither  in  mean  status  score  levels  nor  in  the  relation  between  specific  types 
of  education  and  status  level  in  particular  countries.  Rather,  occupational  status  increases  with  the 
level  of  education  as  well  as  with  the  academic  orientation  of  the  completed  track  at  the  upper 
secondary  level  in  all  European  countries  in  fairly  similar  ways.  In  part,  this  result  reflects  the 
construction  of  the  scale  itself,  which  is  specifically  designed  to  average  out  cross-national  variation  in 
status  evaluations  for  particular  occupations  (cf.  details  in  Ganzeboom  et  al.,  1992,  1996).  Still,  it  is 
remarkable  that  there  is  little  variation  in  the  relation  between  types  of  education  and  status  outcomes 
between  countries.  There  is  much  more  evidence  of  cross-national  variation  in  occupational  allocation 
once  the  discrete  measures  are  considered. 


O 

nr 


5.13 


Figure  1 Unemployment  Rate  by  Country  and  Education 


-Lower  seondary 

-upper  secondary  vocational 

- Lower  tertiary 

-Overall  unemployment  rate 


Apprenticeship 

Upper  secondary  general 

University  degree 


Figure  2 Average  Occupational  Status  (ISEI  scores)  by  Country  and  Education 


"Lower  seondary 
"upper  secondary  vocational 
"Lower  tertiary 
"Average  S tatus 


■Apprenticeship 

"U  pper  secondary  general 

■University  degree 


Sources:  European  Community  Labour  Force  Survey  1992-1997 


3 

tKJC 


5,4  320 


Figure  3 Probability  of  Lower-Skilled  Employment  by  Country  and  Education 


- Lower  seondary 
-upper  secondary  vocational 
• Lower  tertiary 
-Overall  incidence 


Apprenticeship 

Upper  secondary  general 

University  degree 


Figure  4 Probability  of  Professional  Employment  by  Country  and  Education 


-Lower  seondary 
-Upper  secondary  vocational 
-Lower  tertiary 
-Overall  incidence 


Apprenticeship 

Upper  secondary  general 

University  degree 


Sources:  European  Community  Labour  Force  Survey  1992-1997 


O 

ERIC 


AO  I 


5.15 


Turning  to  the  lower  end  of  occupational  outcomes  first,  the  proportion  of  market  entrants  in  lower- 
skilled  employment  - as  defined  above  - varies  substantially  across  countries  and  educational  groups. 
The  overall  proportion  of  market  entrants  in  such  lower-level  occupations  is  lowest  in  Germany  (18%), 
Belgium  (20%),  Austria,  Denmark  and  the  Netherlands  (27%),  ranging  up  to  37%  in  Greece  and  41% 
in  Spain.  Of  course,  it  is  mostly  leavers  with  compulsory  education  only  who  attain  employment  in 
these  occupations,  with  the  incidence  rates  for  this  group  being  at  50%-60%  in  most  countries,  but 
ranging  between  some  40%  in  Italy  and  Portugal  and  up  to  70%  in  Denmark.  In  turn,  leavers  from  both 
levels  of  tertiary  education  are  hardly  found  in  lower-level  jobs.  The  picture  is  less  clear-cut  at  the 
upper  secondary  level,  however.  In  general,  the  incidence  rates  are  at  an  intermediate  level,  but  there 
is  wide  cross-national  variation  in  terms  of  whether  differentiation  between  different  types  of  training  at 
the  upper  secondary  level  exists  at  all,  and  if  so,  which  qualifications  provide  most  favourable 
conditions.  In  most  countries,  it  seems  that  school-based  vocational  education  holds  some  advantages 
over  the  general  academic  tracks,  although  Germany  is  a counter  example  in  that  respect.  Similarly, 
the  relative  status  of  apprenticeship  training  is  not  unequivocal.  While  in  most  countries  apprentices 
perform  equally  well  or  even  better  than  leavers  from  other  upper  secondary  tracks,  this  is  certainly  not 
the  case  in  Austria  or  the  Netherlands,  where  apprentices  are  to  a significant  degree  allocated  to 
lower-level  positions. 

As  a final  indicator,  I will  take  a look  at  the  opposite  pole  of  occupational  hierarchies,  the  level  of 
professional  employment  positions  (cf.  Figure  4).  As  for  lower-level  employment,  there  is  clear 
evidence  of  substantial  cross-national  variation  in  this  discrete  measure  of  occupational  outcomes. 
The  proportion  of  market  entrants  in  these  upper-level  forms  of  employment  ranges  from  about  40%  in 
Belgium  and  Germany,  about  one  third  in  Denmark  and  the  Netherlands  down  to  some  20%  in  Ireland, 
Portugal  and  Spain  and  to  as  low  as  13%  in  Italy.  Evidently,  the  likelihood  of  starting  one's  career  in 
such  positions  is  strongly  related  to  a tertiary  level  educational  background.  Except  for  upper 
secondary  leavers  in  Austria  (and  maybe  Germany),  the  probability  of  entering  these  occupations  is 
virtually  negligible.  This  probability  markedly  increases  for  leavers  from  lower  tertiary  education  and 
improves  even  further  for  university  graduates.  Still,  there  is  substantial  variation  in  these  figures 
between  countries.  The  probability  of  professional  employment  among  lower  tertiary  educated  leavers 
is  mostly  between  45%  and  55%,  but  ranges  between  30%  in  Ireland  and  38%  in  France  up  to  63%  in 
Portugal  and  78%  in  Austria.6  A similar  picture  applies  at  the  level  of  university  graduates:  their 
probability  of  entering  professional  employment  is  at  about  two  thirds  to  70%,  with  Italy  (58%)  and 
Spain  (63%)  deviating  negatively  and  Austria  (88%),  the  Netherlands  (85%)  and  Denmark  (83%) 
providing  particularly  good  prospects. 


In  part  this  might  reflect  a relatively  heterogeneous  coding  of  national  qualifications  to  this  educational  level, 
including  post-secondary  tracks  for  some  countries  and  introducing  questionable  differentiation  between 
tertiary  tracks  for  others  (coding  details  in  Eurostat,  1996). 

' : 5.16 


n n,  /-v 


5 Micro-Macro  Models  of  Labour  Market  Entry 

Having  presented  basic  descriptive  evidence  on  the  relationship  between  education  and  early  labour 
market  outcomes,  the  analyses  now  turn  to  the  multivariate  modelling.  Unemployment  risks  are 
addressed  first,  followed  by  a discussion  of  occupational  attainment  in  terms  of  the  different  indicators 
chosen.  A concluding  section  summarizes  the  implications  of  the  estimated  models  for  explaining 
differences  in  labour  market  entry  patterns  between  institutional  contexts. 


5.1  Unemployment  Risks  at  Entering  the  Labour  Market 

What  explains  unemployment  risks  at  entering  the  labour  market  in  Europe  in  the  mid-1990s?  Details 
on  the  estimated  multilevel  models  are  provided  in  Table  2,  with  estimated  models  successively 
including  additional  variables  to  extract  the  systematic  components  of  country  differences  and 
variation  between  qualifications.  The  estimated  country  level  variance  component  of  the  null  model 
(Model  0)  yields  a substantial  baseline  estimate  on  country  differences  in  unemployment  risks.  Model 
1,  which  includes  random  effects  for  country  and  type  of  education,  shows  that  variation  between 
types  of  education  is  large,  while  variation  between  countries  is  somewhat  reduced,  indicating  that  part 
of  the  cross-national  difference  is  related  to  country  differences  in  the  distribution  of  qualifications 
among  school  leavers.  Both  variance  components  can  be  substantially  reduced  by  subsequent 
models,  introducing  various  systematic  components  into  the  estimation  process.  Taking  account  of  the 
six-category  calssification  of  level  and  type  of  qualifications  in  Model  2 already  accounts  for  about  two 
thirds  of  the  total  variation  of  unemployment  risks  between  types  of  education  (o2=.128  vs.  o2=.320). 
The  parameter  estimates  show  that  unemployment  risks  are  lowest  for  tertiary  qualifications,  and 
lower  for  apprentices  than  for  leavers  from  other  tracks  at  the  lower  or  upper  secondary  level;  a more 
specific  discussion  of  these  and  other  findings  follows  below  where  discrete  change  effects  estimated 
from  the  models  are  discussed.  Besides  the  impact  of  education,  young  women  tend  to  face  higher 
risks  of  unemployment,  as  do  most  recent  entrants  to  the  labour  market.  Models  3 and  4 then  begin  to 
include  country-level  factors  into  the  model.  According  to  Model  3,  including  the  distinction  of  three 
European  institutional  systems,  reduces  the  country-level  variation  present  in  Model  2 by  about  half 
(o2=.241  vs.  o2=.457).  Leavers  in  OLM  systems,  in  general,  face  lowest  unemployment  risk  in  Europe, 
while  unemployment  in  Southern  European  countries  is  well  above  EU  average.  This  conclusion  is 
further  qualified  by  Model  4,  which  includes  economic  context  factors  at  the  country  level.  The  main 
relevant  factor  is  the  countries’  aggregate  unemployment  rate  which  is  strongly  positively  related  to 
market  entrants’  unemployment  risks.  The  findings  for  the  institutional  contexts  remain  qualitatively 
unchanged  after  introducing  this  set  of  controls;  the  changing  magnitudes  indicate  that  part  of  the 
advantageous  performance  of  OLM  systems  in  the  mid-1990s  has  been  due  to  relatively  favourable 
economic  conditions.  Additional  country  variation,  apart  from  the  factors  controlled  for,  appears  to  be 
small  (o2=.04)  as  compared  to  the  initial  estimates  (o2=52).  As  an  addition,  Model  5 establishes  that 
cyclical  movements  of  the  aggregate  unemployment  rate  are  also  the  main  factors  behind  short-term 
changes  in  market  entrants’  unemployment  within  each  country. 


323 


5.17 


Table  2 


Unemployment  Risks  at  Labour  Market  Entry,  Multilevel  Logit  Estimation 


Model  0 
Country 
random 
effect 

Model  1 
Country+ 
education 
random 
effects 

Model  2 
Ml  + 
individual 
factors 

Model  3 
M2  + 
institut. 
systems 

Model  4 
M3  + 
country 
effects 

Model  5 
M4  + 
economic 
trends 

Model  6 
M5  + 

systems  x 
education 

Intercept 

-1.508** 

-1.566“ 

-1.535“ 

-1.507“ 

-1.460“ 

-1.461** 

-1 .467" 

Women 

0.045“ 

0.045** 

0.045** 

0.046" 

0.046" 

1st/2nd  year  leaver 

0.193“ 

0.193“ 

0.193“ 

0.196" 

0.196" 

Education  [Ref:  lower  secondary] 

Apprenticeships 

-0.135 

-0.114 

-0.092 

-0.092 

-0.071 

Upper  secondary  vocational 

0.067 

0.064 

0.043 

0.048 

0.025 

Upper  secondary  general 

0.257“ 

0.250“ 

0.247** 

0.244** 

0.240" 

Lower  tertiary 

-0.319“ 

-0.323* 

-0.317** 

-0.317“ 

-0.299" 

University  degree 

-0.554“ 

-0.558“ 

-0.562“ 

-0.565** 

-0.516" 

Institutional  Systems  [Ref:  ILM] 

OLM  Systems 

-0.709“ 

-0.427“ 

-0.430" 

-0.482“ 

OLM  x Upper  secondary  general 

-0.085 

OLM  x Upper  secondary  vocational 

-0.071 

OLM  x Apprenticeships 

0.055 

OLM  x Lower  tertiary 

-0.291*** 

OLM  x University  degree 

0.057 

Southern  Systems 

0.489“ 

0.656“ 

0.660" 

0.724** 

South  x Upper  secondary  general 

0.075 

South  x Upper  secondary  vocational 

-0.027 

South  x Lower  tertiary 

0.495** 

South  x University  degree 

0.117 

Between-Country  Context 

Unemployment  Rate 

0.152“ 

0.152** 

0.150“ 

Youth-Adult  Ratio 

-0.062 

-0.063 

-0.063 

% Tertiary  Level  Qualifications 

-0.002 

-0.002 

-0.001 

% Professional  Employment 

0.040 

0.040 

0.041 

Within-Country  Changes 

Business  Cycle 

0.102“ 

0.102" 

Youth-Adult  Ratio 

0.004 

0.004 

Educational  Expansion 

0.019 

0.020 

Occupational  Upgrading 

0.012 

0.013 

Variance  Components 

Cf2Education 

i 

0.320 

0.128 

0.127 

0.130 

0.124 

0.061 

Cf2Country 

0.519 

0.438 

0.457 

0.241 

0.035 

0.036 

0.043 

Deviance 

79,232 

76,445 

76,023 

76,023 

76,022 

75,840 

75,842 

N 

78,955 

78,955 

78,955 

78,955 

78,955 

78,955 

78,955 

Periods 

66 

66 

66 

66 

66 

66 

66 

Educational  Groups 

/ 

63 

63 

63 

63 

63 

63 

Countries 

12 

12 

12 

12 

12 

12 

12 

Notes:  Statistical  significance  at  "p<.05  and  *p<.10;  all  variables  are  entered  effect-coded  and  mean-centered; 
(n.e.)  - country-level  random  effect  not  estimated  due  to  convergence  problems. 


Source:  European  Community  Labour  Force  Survey  1992-1997 


O 

ERIC 


5.18 


Table  3 


Unemployment  Risks  at  Labour  Market  Entry,  Discrete  Change  Effects 


Overall 

Lower 

secondary 

Educational  Qualification 
Apprentice  Upper  Upper 

-ships  secondary  secondary 
vocational  general 

Lower 

tertiary 

University 

degree 

Average  Prediction 

0.202 

0.300 

0.130 

0.191 

0.227 

0.146 

0.121 

Institutional  Systems 
ILM  Systems 

-0.014 

+0.018 

+0.007 

-0.021 

-0.038 

-0.047 

-0.038 

OLM  Systems 

-0.046 

-0.030 

-0.007 

-0.071 

-0.084 

-0.073 

-0.038 

Southern  Europe 

+0.115 

+0.014 

N/A 

+0.131 

+0.168 

+0.221 

+0.121 

Between-Country  Economic  Factors 
Aggregate  Unemployment  Rate  +0.024 

+0.032 

+0.018 

+0.024 

+0.027 

+0.020 

+0.017 

Youth-Adult  Ratio 

-0.010 

-0.013 

-0.007 

-0.009 

-0.011 

-0.008 

-0.006 

% Tertiary  Level  Qualifications 

+0.000 

+0.000 

+0.000 

+0.000 

+0.000 

+0.000 

+0.000 

% Professional  Employment 

+0.007 

+0.009 

+0.005 

+0.006 

+0.007 

+0.005 

+0.004 

Within-Country  Economic  Trends 
Business  Cycle 

+0.016 

+0.022 

+0.012 

+0.016 

+0.018 

+0.013 

+0.011 

Youth-Adult  Ratio 

+0.001 

+0.001 

+0.000 

+0.001 

+0.001 

+0.000 

+0.000 

Educational  Expansion 

+0.003 

+0.004 

+0.002 

+0.003 

+0.003 

+0.002 

+0.002 

Occupational  Upgrading 

+0.002 

+0.003 

+0.001 

+0.002 

+0.002 

+0.002 

+0.001 

Notes:  Predicted  probabilities  based  on  final  model  at  the  means  of  covariates;  figures  represent  discrete 

change  effects  for  dummy  variables  and  unit  effects  of  covariates  respectively.  Cf.  Table  1 for  further 
details. 


Figure  5 Predicted  Unemployment  Probabilities  by  Institutional  Contexts  and  Education 


Lower  secondary  Apprenticeships  Upper  secondary  Upper  secondary  Lower  tertiary  University  degree 

vocational  general 


Notes:  Predicted  probabilities  based  on  final  model  at  the  means  of  covariates. 


O 

ERIC 


5.19 


325 


Model  6,  finally,  includes  interaction  effects  between  qualifications  and  institutional  systems,  allowing 
the  role  of  education  to  vary  between  contexts.  As  a result,  variation  between  types  of  education  is 
reduced  to  o2=.06,7  mostly  due  to  a relatively  better  standing  of  leavers  from  lower  tertiary  education 
and  relatively  larger  unemployment  risks  of  the  lowest  qualified  in  OLM  contexts,  while  the  reverse 
pattern  holds  for  Southern  Europe.8 

Substantive  interpretation  of  complex  and  non-linear  models  like  the  ones  estimated  is  much  facilitated 
by  means  of  providing  predicted  probabilities  for  a core  set  of  features  of  interest,  allowing  us  to  add  a 
more  quantitative  interpretation  of  estimated  effects  to  the  qualitative  one  already  provided  above. 
Table  3 provides  these  predictions  both  at  the  aggregate  level  and  by  specific  type  of  education  for  all 
context-level  variables,  each  prediction  calculated  at  the  means  of  all  other  covariates.  Figure  1 
complements  the  presentation  by  providing  a graphical  display  of  educational  effects  in  each  of  the 
three  institutional  systems.  According  to  these  results,  leavers  from  tertiary  education  and 
apprenticeship  training  face  the  most  favourable  prospects  at  entering  the  labour  market,  with 
predicted  probabilities  of  unemployment  between  .121  and  .146,  or  equivalently,  .06  to  .08  below  the 
average  prediction.  The  probability  of  unemployment  for  leavers  from  school-based  upper  secondary 
education  is  around  the  overall  average  of  .20,  with  a relative  advantage  of  .03  to  vocationally  qualified 
leavers.  It  is  the  least  qualified  who,  finally,  face  the  highest  unemployment  probabilities  of  about  .10 
above  the  average  figure,  or  .300  in  total.  In  addition,  these  relations  differ  markedly  between 
institutional  contexts,  implying  a probability  differential  of  .16  between  unemployment  risks  in  OLM 
systems  versus  Southern  European  systems  (-.046  to  +.115)  solely  due  to  systemic  differences  in  the 
evaluation  of  otherwise  identical  qualifications.  More  specifically,  unemployment  risks  for  all  but 
leavers  with  compulsory  education  only  are,  at  otherwise  identical  economic  context,  at  least  .10 
higher  in  Southern  Europe,  leading  to  the  well-known  bell-shaped  rather  than  negative  relationship 
between  education  and  unemployment  (OECD  1996,  1997).  In  addition,  unemployment  risks  in  OLM 
systems  are  also  substantially  lower  than  those  for  the  equivalent  qualificational  groups  in  ILM 
contexts,  notably  among  the  lowest  qualified  and  at  the  level  of  upper  secondary  and  lower  tertiary 
qualifications.  The  predictions  finally  also  stress  the  crucial  role  of  aggregate  labour  market  conditions 
for  market  entrants’  unemployment  risks.  Comparing  the  unit  effect  both  between  and  within  countries 
emphasizes  that  aggregate  labour  market  conditions  disproportionally  affect  unemployment  risks  of 
market  entrants:  the  estimates  amount  to  predicting  an  overall  increase  in  market  entrants’ 
unemployment  of  2.4%  associated  with  a 1%  increase  in  a countries’  average  aggregate 
unemployment  rate  over  the  1992-1997  period,  and  a 1.6%  increase  following  a 1%  deterioration  of 
the  aggregate  unemployment  rate  within  any  country.  The  apparent  relation  between  the  magnitude  of 
estimated  effects  and  the  level  of  unemployment  risks  for  any  educational  group  is,  of  course,  a 
consequence  of  the  non-linear  model  specification  adopted.  Both  the  substantive  issue  of  interactions 


The  variance  estimate  for  type  of  education  in  the  null  model  (without  nesting  within  country)  is  aJ=.72. 

In  addition  to  the  estimation  results  provided,  separate  models  for  the  more  recent  and  more  experienced 
market  entrants  have  been  fitted  for  all  dependent  variables  in  order  to  explore  the  extent  to  which  core  results 
might  be  systematically  more  pronounced  among  the  more  recent  entrants.  There  are,  however,  hardly  any 
indications  that  pooling  the  first  five  years  in  the  labour  market'affieptethe  results  in  any  meaningful  way. 


£ 05.20 


between  trends  and  type  of  qualification  as  well  as  the  question  of  interactions  between  institutional 
systems  and  trend  factors  are  not  pursued  any  further  here  (but  cf.  Gangl,  2000b). 

5.2  Occupational  Allocation  of  Market  Entrants 

Turning  from  unemployment  to  employment  outcomes,  the  analyses  now  address  the  occupational 
allocation  of  market  entrants  across  EU  countries.  In  doing  so,  the  paper  relies  on  a multiple  rather 
than  singular  characterization  of  occupational  outcomes,  drawing  both  on  a continuously  scaled 
measure  of  occupational  status  and  two  discrete  measures  for  both  the  most  and  least  advantageous 
occupations,  respectively.  The  presentation  of  results  will,  in  general,  follow  the  format  established  in 
the  case  of  unemployment.  In  more  substantive  terms,  the  most  general  measure  of  occupational 
status  will  be  discussed  first,  while  results  on  both  the  incidence  on  lower-skilled  employment,  and  the 
attainment  of  professional  employment  positions  follow  subsequently. 

5.2.1  Status  Attainment 

Occupational  status  attainment  is  measured  according  to  the  ISEI  scale  here  (cf.  Ganzeboom  et  al., 
1992,  1996)  and  results  from  the  multilevel  regressions  of  these  scores  are  presented  in  Table  4,  with 
predicted  status  scores  and  discrete  change  effects  again  being  given  in  Table  5 and  Figure  2.  In 
general,  variation  between  countries  in  terms  of  ISEI  scores  is  small  (a2=.003  on  the  log-normal 
scale),  and  mostly  related  to  cross-national  differences  in  the  distribution  of  qualifications  among 
market  entrants  rather  than  any  country-level  context  factor  controlled  for  in  the  regression.  From  the 
set  of  between-country  context  factors,  only  the  level  of  tertiary  level  graduates  in  the  labour  force  is 
weakly  negatively  related  to  occupational  outcomes.  In  a clearer  fashion,  educational  expansion  and 
occupational  upgrading  in  the  labour  market  within  each  country  contribute  (in  counteracting  ways)  to 
short-term  changes  in  occupational  allocation,  with  educational  expansion  leading  to  reduced,  and 
occupational  upgrading  implying  increasing,  levels  of  status  attainment.  Increasing  youth-adult  ratios 
in  the  market  also  correspond  with  small  increases  in  status  attainment. 

In  turn,  status  attainment  is  more  clearly  linked  to  individual  resources.  Young  women  tend  to  achieve 
slightly  more  favourable  employment  positions,  while  the  reverse  result  holds  for  the  most  recent 
entrants  to  the  market.  The  role  of  education  is  crucial,  however,  and  accounting  for  the  level  and  type 
of  qualifications  taps  most  of  the  existing  variation  between  qualifications  in  terms  of  average 
occupational  status  outcomes.  At  average  context  conditions,  the  model  predicts  a status  score  of 
31.7  for  the  lowest  qualified,  with  returns  to  apprenticeships  of  about  3.5  status  points,  7 points  for 
leavers  from  upper  secondary  vocational  education,  8.5  status  points  for  entrants  from  upper 
secondary  general  education,  16  ISEI  score  points  for  leavers  from  lower  tertiary  education  and  up  to 
30  points  for  university  graduates.  In  addition,  there  is  no  evidence  of  strong  interactions  between 
educational  effects  and  institutional  systems,  so  that  leavers  in  OLM  contexts  perform  only  slightly 
better  overall  than  do  their  counterparts  in  both  ILM  and  Southern  systems. 


Table  4 Status  Attainment  at  Labour  Market  Entry,  Multilevel  Estimation  (Log  ISEI  Score) 


Model  0 
Country 
random 
effect 

Model  1 
Country* 
education 
random 
effects 

Model  2 
Ml  + 
individual 
factors 

Model  3 
M2  + 
institut. 
systems 

Model  4 
M3  + 
country 
effects 

Model  5 
M4  + 
economic 
trends 

Model  6 
M5  + 

systems  x 
education 

Intercept 

3.757* 

3.750“ 

3.727“ 

3.731“ 

3.726“ 

3.726" 

3.726" 

Women 

0.012“ 

0.012“ 

0.012“ 

0.012" 

0.012" 

1st/2nd  year  leaver 

-0.018“ 

-0.018“ 

-0.018“ 

-0.018" 

-0.018** 

Education  [Ref:  lower  secondary] 

Apprenticeships 

-0.168“ 

-0.169“ 

-0.171“ 

-0.171** 

-0.172** 

Upper  secondary  vocational 

-0.073“ 

-0.071“ 

-0.072“ 

-0.073" 

-0.072" 

Upper  secondary  general 

-0.037“ 

-0.037“ 

-0.035* 

-0.035* 

-0.030 

Lower  tertiary 

0.140“ 

0.139“ 

0.140“ 

0.140" 

0.138** 

University  degree 

0.401“ 

0.401“ 

0.402“ 

0.402" 

0.405** 

Institutional  Systems  [Ref:  ILM] 

OLM  Systems 

0.019 

0.021 

0.021 

0.020 

OLM  x Upper  secondary  general 

-0.009 

OLM  x Upper  secondary  vocational 

-0.010 

OLM  x Apprenticeships 

0.001  . 

OLM  x Lower  tertiary 

0.015 

OLM  x University  degree 

0.017 

Southern  Systems 

0.016 

-0.013 

-0.013 

-0.015 

South  x Upper  secondary  general 

0.038 

South  x Upper  secondary  vocational 

0.016 

South  x Lower  tertiary 

-0.031 

South  x University  degree 

0.008 

Between-Country  Context 

Unemployment  Rate 

-0.002 

-0.002 

-0.002 

Youth-Adult  Ratio 

0.002 

0.002 

0.002 

% Tertiary  Level  Qualifications 

-0.007* 

-0.007* 

-0.007* 

% Professional  Employment 

0.002 

0.002 

0.002 

Within-Country  Changes 

Business  Cycle 

0.001 

0.001 

Youth-Adult  Ratio 

0.001" 

o 

o 

o 

— k 

Educational  Expansion 

-0.014“ 

-0.014“ 

Occupational  Upgrading 

0.021“ 

0.021" 

Variance  Components 

^Education 

/ 

0.058 

0.004 

0.004 

0.004 

0.004 

0.004 

^Country 

0.003 

(n.e.) 

0.002 

0.001 

0.001 

0.001 

0.001 

Deviance 

59,294 

59,242 

59,244 

59,244 

59,244 

59,240 

59,239 

N 

59,306 

59,306 

59,306 

59,306 

59,306 

59,306 

59,306 

Periods 

66 

66 

66 

66 

66 

66 

66 

Educational  Groups 

/ 

63 

63 

63 

63 

63 

63 

Countries 

12 

12 

12 

12 

12 

12 

12 

Notes:  Statistical  significance  at  **p<.05  and  *p<.1 0;  all  variables  are  entered  effect-coded  and  mean-centered; 
(n.e.)  - country-level  random  effect  not  estimated  due  to  convergence  problems. 


Source:  European  Community  Labour  Force  Survey  1992-1997 


O 

ERIC 


•;»  C 


5.22 


Table  5 Status  Attainment  at  Labour  Market  Entry,  Discrete  Change  Effects  (ISEI  Score) 


Overall 

Lower 

secondary 

Educational  Qualification 
Apprentice  Upper  Upper 

-ships  secondary  secondary 
vocational  general 

Lower 

tertiary 

University 

degree 

Average  Prediction 

41.4 

31.7 

35.2 

38.6 

40.3 

47.6 

62.2 

Institutional  Systems 
ILM  Systems 

-0.28 

1.28 

-0.47 

-0.45 

-1.35 

+0.50 

-1.80 

OLM  Systems 

+0.80 

+0.17 

+0.47 

+0.40 

+0.46 

+1.70 

+2.32 

Southern  Europe 

-0.59 

-1.39 

N/A 

+0.05 

+0.93 

-2.12 

-0.45 

Between-Country  Economic  Factors 
Aggregate  Unemployment  Rate 

-0.10 

-0.08 

-0.08 

-0.09 

-0.10 

-0.11 

-0.15 

Youth-Adult  Ratio 

+0.08 

+0.06 

+0.07 

+0.07 

+0.08 

+0.09 

+0.12 

% Tertiary  Level  Qualifications 

-0.30 

-0.23 

-0.25 

-0.28 

-0.29 

-0.34 

-0.45 

% Professional  Employment 

+0.07 

+0.05 

+0.06 

+0.07 

+0.07 

+0.08 

+0.11 

Within-Country  Economic  Trends 
Business  Cycle 

+0.03 

+0.03 

+0.03 

+0.03 

+0.03 

+0.04 

+0.05 

Youth-Adult  Ratio 

+0.06 

+0.04 

+0.05 

+0.05 

+0.06 

+0.07 

+0.09 

Educational  Expansion 

-0.59 

-0.45 

-0.50 

-0.55 

-0.58 

-0.68 

-0.89 

Occupational  Upgrading 

+0.89 

+0.68 

+0.75 

+0.83 

+0.86 

+1.02 

+1.33 

Notes:  Predictions  based  on  final  model  at  the  means  of  covariates;  figures  represent  discrete  change  effects 

for  dummy  variables  and  unit  effects  of  covariates  respectively.  Cf.  Table  3 for  further  details. 


Figure  6 Predicted  ISEI  Scores  by  Institutional  Contexts  and  Education 


Notes:  Predicted  status  scores  based  on  final  model  at  the  means  of  covariates. 


‘ 5.23 


329 


5.2.2  Incidence  of  Lower-Skilled  Employment 

As  status  scores  provide  summary  measures  of  different  aspects  of  employment  positions,  they  are 
notoriously  difficult  to  interpret.  Therefore,  two  alternative,  discrete  measures  of  occupational 
allocation  are  considered  here,  namely  allocation  at  the  top  and  bottom  level  of  the  occupational 
hierarchy.  Turning  to  the  latter  first,  Table  6 presents  the  outcomes  of  estimated  multilevel  models  of 
the  incidence  of  such  lower-level  employment  at  market  entry.  Again,  Table  7 gives  the  estimates  of 
discrete  change  effects  and  Figure  3 graphically  depicts  the  predicted  probabilities  by  educational 
groups  and  institutional  systems.  Compared  to  the  case  of  unemployment,  overall  cross-national 
variation  in  this  indicator  is  much  smaller  (o2=.1 1 in  Model  0 vs.  o2=.52),  yet  variation  between  types  of 
qualifications  much  more  important  (o2=1.48  in  Model  1 vs.  o2=.72).  Most  of  this  variation  is  again 
captured  by  differentiating  between  the  six  types  of  qualifications,  with  little  evidence  of  strong 
interactions  between  educational  effects  and  institutional  contexts.  Nor  is  there  evidence  of  substantial 
differences  in  the  incidence  of  lower-skilled  employment  between  institutional  systems  or  according  to 
measured  economic  context  factors.  It  is  only  that  increasing  educational  expansion  within  countries 
fosters,  while  increasing  occupational  upgrading  appears  to  reduce,  the  probability  of  lower-level 
employment  at  entering  the  labour  market  in  the  short  run.  Actually,  most  of  the  country  differences 
explained  by  the  introduction  of  systematic  factors  into  the  model  are  related  to  cross-national 
differences  in  educational  distributions,  leaving  a still  relatively  large  proportion  of  overall  country 
differences  unexplained. 

With  respect  to  individual-level  factors,  recent  market  entrants  again  face  relatively  less  favourable 
outcomes,  while  the  evidence  on  gender  differences  is  weak  at  best.  However,  as  indicated  above, 
educational  differences  are  pronounced.  In  average  context  conditions,  the  probability  of  being  in  any 
kind  of  lower-level  occupation  within  the  first  five  years  in  the  market  is  estimated  at  .55  for  the  lowest 
qualified,  but  at  .045  only  for  university  leavers.  The  predicted  probability  for  leavers  from  lower  tertiary 
education  is  slightly  less  than  .13,  while  the  respective  figures  for  leavers  from  the  different  types  of 
upper  secondary  education  and  training  are  around  one  third.  There  is  slight  evidence  of  institutional 
variation  in  the  role  of  education  here,  with  educational  stratification  being  slightly  more  pronounced  in 
OLM  systems:  there,  the  lowest  qualified  leavers  have  somewhat  higher  risk  of  ending  up  in  lower- 
level  jobs,  while  the  reverse  is  true  for  school-based  upper  secondary  and  tertiary  qualifications. 
Systematic  differences  between  ILM  and  Southern  countries  are  hardly  discernible. 


5.24 


330 


Table  6 


Lower-Skilled  Employment  at  Labour  Market  Entry,  Multilevel  Logit  Estimation 


Model  0 
Country 
random 
effect 

Model  1 
Country* 
education 
random 
effects 

Model  2 
Ml  + 
individual 
factors 

Model  3 
M2  + 
institut. 
systems 

Model  4 
M3  + 
country 
effects 

Model  5 
M4  + 
economic 
trends 

Model  6 
M5  + 

systems  x 
education 

Intercept 

-0.890" 

-1.212" 

-1.100" 

-1.123" 

o> 

o> 

p 

T — 
1 

-1.104" 

-1.117" 

Women 

0.017* 

0.017* 

0.017* 

0.018* 

0.018* 

1st/2nd  year  leaver 

0.110" 

0.110" 

0.110" 

0.108" 

0.108" 

Education  [Ref:  lower  secondary] 

Apprenticeships 

0.507" 

0.521** 

0.540" 

0.539** 

0.493" 

Upper  secondary  vocational 

0.389" 

0.379" 

0.383" 

0.386** 

0.396** 

Upper  secondary  general 

0.553" 

0.549" 

0.540" 

0.541" 

0.542" 

Lower  tertiary 

-0.821" 

-0.816" 

-0.822" 

-0.834" 

-0.811" 

University  degree 

-1.926" 

-1.932" 

-1.939" 

-1 .943" 

-1.941" 

Institutional  Systems  [Ref:  ILM] 

OLM  Systems 

-0.170 

-0.117 

-0.118 

-0.151 

OLM  x Upper  secondary  general 

-0.079 

OLM  x Upper  secondary  vocational 

-0.005 

OLM  x Apprenticeships 

0.261* 

OLM  x Lower  tertiary 

-0.218 

OLM  x University  degree 

• 

o 

k> 

CO 

k 

Southern  Systems 

-0.059 

0.044 

0.043 

0.089 

South  x Upper  secondary  general 

-0.005 

South  x Upper  secondary  vocational 

-0.004 

South  x Lower  tertiary 

0.089 

South  x University  degree 

0.174 

Between-Country  Context 

Unemployment  Rate 

0.025 

0.025 

0.026 

Youth-Adult  Ratio 

0.020 

0.019 

0.009 

% Tertiary  Level  Qualifications 

0.020 

0.021 

0.024 

% Professional  Employment 

-0.003 

-0.004 

-0.003 

Within-Country  Changes 

Business  Cycle 

-0.005 

-0.005 

Youth-Adult  Ratio 

0.002 

0.002 

Educational  Expansion 

o 

k 

o 

a>. 

0.106" 

Occupational  Upgrading 

-0.059* 

-0.059* 

Variance  Components 

(^Education 

/ 

1.484 

0.121 

0.121 

0.122 

0.125 

0.124 

^Country 

0.113 

(n.e.) 

0.062 

0.040 

0.049 

0.047 

0.046 

Deviance 

73,343 

63,869 

63,779 

63,779 

63,778 

63,682 

63,678 

N 

63,715 

63,715 

63,715 

63,715 

63,715 

63,715 

63,715 

Periods 

66 

66 

66 

66 

66 

66 

66 

Educational  Groups 

/ 

63 

63 

63 

63 

63 

63 

Countries 

12 

12 

12 

12 

12 

12 

12 

Notes:  Statistical  significance  at  p<.05  and  p<.10;  all  variables  are  entered  effect-coded  and  mean-centered, 
(n.e.)  - country-level  random  effect  not  estimated  due  to  convergence  problems 


Source:  European  Community  Labour  Force  Survey  1992-1997 


5.25 


331 


Table  7 


Lower-Skilled  Employment  at  Labour  Market  Entry,  Discrete  Change  Effects 


Overall 

Lower 

secondary 

Educational  Qualification 
Apprentice  Upper  Upper 

-ships  secondary  secondary 
vocational  general 

Lower 

tertiary 

University 

degree 

Average  Prediction 

0.329 

0.551 

0.339 

0.327 

0.360 

0.127 

0.045 

Institutional  Systems 
ILM  Systems 

+0.005 

-0.004 

-0.034 

+0.016 

+0.034 

+0.023 

+0.008 

OLM  Systems 

-0.001 

+0.044 

+0.035 

-0.033 

-0.051 

-0.036 

-0.016 

Southern  Europe 

+0.003 

-0.041 

N/A 

+0.019 

+0.020 

+0.021 

+0.013 

Between-Country  Economic  Factors 
Aggregate  Unemployment  Rate  +0.005 

+0.006 

+0.006 

+0.006 

+0.006 

+0.003 

+0.001 

Youth-Adult  Ratio 

+0.002 

+0.002 

+0.002 

+0.002 

+0.002 

+0.001 

+0.000 

% Tertiary  Level  Qualifications 

+0.005 

+0.006 

+0.005 

+0.005 

+0.006 

+0.003 

+0.001 

% Professional  Employment 

-0.001 

-0.001 

-0.001 

-0.001 

-0.001 

-0.000 

-0.000 

Within-Country  Economic  Trends 
Business  Cycle 

-0.001 

-0.001 

-0.001 

-0.001 

-0.001 

-0.001 

+0.000 

Youth-Adult  Ratio 

+0.000 

+0.001 

+0.001 

+0.001 

+0.001 

+0.000 

+0.000 

Educational  Expansion 

+0.020 

+0.026 

+0.024 

+0.024 

+0.025 

+0.012 

+0.005 

Occupational  Upgrading 

-0.011 

-0.015 

-0.013 

-0.013 

-0.013 

-0.006 

-0.002 

Notes:  Predicted  probabilities  based  on  final  model  at  the  means  of  covariates;  figures  represent  discrete 
change  effects  for  dummy  variables  and  unit  effects  of  covariates  respectively.  Cf.  Table  5 for  further 
details. 


Figure  7 Predicted  Probabilities  by  institutional  Contexts  and  Education 


Lower  secondary  Apprenticeships  Upper  secondary  Upper  secondary  Lower  tertiary  University  degree 

vocational  general 


Notes:  Predicted  probabilities  based  on  final  model  at  the  means  of  covariates. 


o 

ERIC 


5.26  332 


5.2.3  Access  to  Professional  Employment  Positions 

As  a final  indicator  of  initial  occupational  allocation,  the  probability  of  access  to  professional 
employment  positions  at  entering  the  market  is  considered.  And  as  before,  Table  8 provides  full  details 
of  the  estimated  models,  while  Table  9 holds  the  predicted  probabilities  and  discrete  change  effects, 
which  are  in  part  depicted  graphically  in  Figure  4.  The  two  null  models  (Models  0/1)  indicate 
substantial  variation  in  this  variable,  both  cross-nationally  (o2=.204)  and  between  types  of  education 
(o2=3.267),  which  is  consistently  larger  than  with  respect  to  the  incidence  of  lower-skilled  employment. 
And  as  in  the  previous  estimation,  cross-national  variation  in  terms  of  occupational  allocation  is  much 
smaller  than  was  the  case  for  unemployment,  while  variation  between  types  of  education  is 
considerably  larger.  But  again,  most  of  the  variance  between  types  of  education  is  accounted  for  by 
differentiating  the  six  measured  categories  of  education.  There  is  almost  no  evidence  of  a systematic 
interaction  between  educational  effects  and  institutional  systems.  But  there  is  evidence  of  an 
institutional  difference  at  the  country  level,  reducing  the  country  level  variation  by  more  than  half 
(o2=.11  in  Model  3 vs.  o2=.26  in  Model  2).  According  to  the  parameter  estimates  in  Model  3,  market 
entrants  in  OLM  systems  have  a consistently  higher  probability  of  professional  employment  than  do 
their  counterparts  in  both  Southern  and  ILM  systems.  As  the  estimation  provided  by  Model  4 clarifies, 
part  of  this  favourable  situation  is  due  to  a more  favourable  aggregate  labour  market  situation  and,  on 
average,  a smaller  proportion  of  tertiary  qualifications  in  the  market,  both  of  which  notably  improve  the 
relative  position  of  OLM  and  Southern  Europe  versus  ILM  systems.  Still,  a systematically  more 
favourable  situation  prevails  in  OLM  systems  after  controlling  for  country-level  economic  context 
factors,  which  captures  most  of  the  variance  between  countries.  In  addition,  the  same  trend  effects  as 
for  status  attainment  can  be  established,  with  the  probability  of  attaining  professional  positions 
declining  with  educational  expansion,  increasing  with  occupational  upgrading,  and  weakly  increasing 
with  growing  youth-adult  ratios  in  the  labour  force. 

Turning  to  the  individual-level  factors,  I find  a negative  effect  on  the  probability  of  professional 
employment  for  women  and  most  recent  market  entrants.  The  pervasive  effects  of  educational 
qualifications  are  again  best  described  by  considering  the  predicted  probabilities  at  the  average 
context  conditions,  as  provided  in  Table  9 and  Figure  4,  respectively.  Of  course,  attainment  of  such 
prestigious  positions  is  basically  confined  to  leavers  from  tertiary  education,  with  predicted 
probabilities  for  secondary-level  leavers  not  exceeding  a level  of  about  .10  in  any  case.  For  tertiary 
level  leavers,  the  model  estimation  predicts  an  average  probability  of  .45  for  leavers  from  lower  tertiary 
education,  and  .71  for  university  graduates.  In  this  case,  there  is  also  evidence  of  variation  between 
institutional  systems,  with  relatively  larger  advantages  to  both  types  of  tertiary  qualifications  in  OLM 
contexts  and  to  lower  tertiary  leavers  in  Southern  Europe.  The  relatively  more  advantageous  position 
of  tertiary  leavers  in  OLM  contexts  leads  to  a probability  differential  of  about  .05  as  compared  to  the 
average  prediction  for  both  groups  of  tertiary  leavers,  and  a differential  of  .10  between  OLM  systems 
and  the  least  advantageous  system,  respectively  (ILM  systems  for  lower  tertiary,  and  Southern  Europe 
for  university  graduates). 


5.27 


333 


Table  8 Professional  Employment  at  Labour  Market  Entry,  Multilevel  Logit  Estimation 


Model  0 
Country 
random 
effect 

Model  1 
Country + 
education 
random 
effects 

Mode!  2 
Ml  + 
individual 
factors 

Model  3 
M2  + 
institut. 
systems 

Model  4 
M3  + 
country 
effects 

Model  5 
M4  + 
economic 
trends 

Model  6 
M5  + 

systems  x 
education 

Intercept 

-1.076" 

-1.572“ 

-1.763“ 

-1.735“ 

-1 .783“ 

-1.786" 

-1.785“ 

Women 

-0.035“ 

-0.035“ 

-0.035“ 

-0.036“ 

-0.036" 

1st/2nd  year  leaver 

-0.173“ 

-0.173“ 

-0.173“ 

-0.173“ 

-0.173“ 

Education  [Ref:  lower  secondary] 

Apprenticeships 

-1.421“ 

-1.435“ 

-1.446“ 

-1.449" 

-1.437** 

Upper  secondary  vocational 

-0.404“ 

-0.393“ 

-0.407“ 

-0.413“ 

-0.404" 

Upper  secondary  general 

-0.499“ 

-0.496“ 

-0.483“ 

-0.482" 

-0.467“ 

Lower  tertiary 

1.557“ 

1 .549“ 

1 .554” 

1.563" 

1.597“ 

University  degree 

2.658“ 

2.662“ 

2.668“ 

2.677“ 

2.683“ 

Institutional  Systems  [Ref:  ILM] 

OLM  Systems 

0.504“ 

0.352” 

0.354" 

0.370“ 

OLM  x Upper  secondary  general 

-0.070 

OLM  x Upper  secondary  vocational 

0.005 

OLM  x Apprenticeships 

-0.061 

OLM  x Lower  tertiary 

-0.171 

OLM  x University  degree 

-0.075 

Southern  Systems 

-0.082 

-0.263 

-0.266 

-0.279 

South  x Upper  secondary  general 

0.165 

South  x Upper  secondary  vocational 

0.025 

South  x Lower  tertiary 

0.344 

South  x University  degree 

0.082 

Between-Country  Context 

Unemployment  Rate 

-0.058* 

-0.058* 

-0.054 

Youth-Adult  Ratio 

-0.040 

-0.039 

-0.033 

% Tertiary  Level  Qualifications 

-0.040* 

-0.040* 

-0.044* 

% Professional  Employment 

0.009 

0.009 

0.013 

Within-Country  Changes 

Business  Cycle 

0.008 

0.007 

Youth-Adult  Ratio 

0.010“ 

0.010” 

Educational  Expansion 

-0.111“ 

-0.111*’ 

Occupational  Upgrading 

0.168“ 

0.169“ 

Variance  Components 

^Education 

/ 

3.267 

0.168 

0.165 

0.167 

0.173 

0.168 

^Country 

0.204 

(n.e.) 

0.257 

0.111 

0.020 

0.020 

0.029 

Deviance 

71,377 

47,947 

47,784 

47,785 

47,784 

47,745 

47,742 

N 

63,742 

63,742 

63,742 

63,742 

63,742 

63,742 

63,742 

Periods 

66 

66 

66 

66 

66 

66 

66 

Educational  Groups 

/ 

63 

63 

63 

63 

63 

63 

Countries 

12 

12 

12 

12 

12 

12 

12 

Notes:  Statistical  significance  at  **p<.05  and  *p<.10;  all  variables  are  entered  effect-coded  and  mean-centered, 
(n.e.)  - country-level  random  effect  not  estimated  due  to  convergence  problems 


Source:  European  Community  Labour  Force  Survey  1992-1997 


ERIC 


5.28 


Table  9 Professional  Employment  at  Labour  Market  Entry,  Discrete  Change  Effects 


Overall 

Lower 

secondary 

Educational  Qualification 
Apprentice  Upper  Upper 

-ships  secondary  secondary 
vocational  general 

Lower 

tertiary 

University 

degree 

Average  Prediction 

0.208 

0.023 

0.044 

0.101 

0.095 

0.453 

0.711 

Institutional  Systems 
ILM  Systems 

-0.012 

+0.004 

-0.007 

-0.010 

-0.015 

-0.064 

-0.021 

OLM  Systems 

+0.033 

+0.024 

+0.008 

+0.039 

+0.029 

+0.050 

+0.057 

Southern  Europe 

-0.018 

-0.013 

N/A 

-0.021 

-0.009 

+0.016 

-0.042 

Between-Country  Economic  Factors 
Aggregate  Unemployment  Rate  -0.005 

-0.001 

-0.002 

-0.005 

-0.005 

-0.013 

-0.011 

Youth-Adult  Ratio 

-0.003 

-0.001 

-0.001 

-0.003 

-0.003 

-0.008 

-0.007 

% Tertiary  Level  Qualifications 

-0.004 

-0.001 

-0.002 

-0.004 

-0.004 

-0.011 

-0.009 

% Professional  Employment 

+0.001 

+0.000 

+0.001 

+0.001 

+0.001 

+0.003 

+0.003 

Within-Country  Economic  Trends 
Business  Cycle 

+0.001 

+0.000 

+0.000 

+0.001 

+0.001 

+0.002 

+0.001 

Youth-Adult  Ratio 

+0.001 

+0.000 

+0.000 

+0.001 

+0.001 

+0.002 

+0.002 

Educational  Expansion 

-0.011 

-0.002 

-0.004 

-0.010 

-0.009 

-0.027 

-0.023  . 

Occupational  Upgrading 

+0.017 

+0.004 

+0.008 

+0.016 

+0.016 

+0.042 

+0.033 

Notes:  Predicted  probabilities  based  on  final  model  at  the  means  of  covariates;  figures  represent  discrete 
change  effects  for  dummy  variables  and  unit  effects  of  covariates  respectively.  Cf.  Table  7 for  further 
details. 


Figure  8 Predicted  Probabilities  by  Institutional  Contexts  and  Education 


Lower  secondary  Apprenticeships  Upper  secondary  Upper  secondary  Lower  tertiary  University  degree 

vocational  general 


Notes:  Predicted  probabilities  based  on  final  model  at  the  means  of  covariates. 


5.29 


335 


5.3  Comparing  Labour  Market  Entry  across  Institutional  Systems 

Having  performed  a set  of  comparative  analyses  of  unemployment  and  occupational  allocation  at 
labour  market  entry,  where  the  role  of  educational  resources  and  the  role  of  different  institutional 
contexts  for  market  entry,  both  providing  different  qualificational  backgrounds  to  market  entrants  and 
varying  conditions  for  converting  educational  resources  into  labour  market  outcomes,  has  been 
emphasized,  the  final  question  to  be  addressed  is  whether  and  to  what  extent  such  institutional 
differences  are  able  to  explain  the  observed  differences  between  institutional  systems  in  terms  of 
labour  market  entry  patterns.  The  results  of  a final  analysis  decomposing  the  empirical  differences 
between  the  three  institutional  systems  differentiated  throughout  the  paper  according  to  the  systematic 
effects  established  in  the  multivariate  analyses  are  given  in  Table  10.  There,  empirical  differences  in 
aggregate  labour  market  entry  outcomes  between  the  systems  are  related  to  system  level  differences 
in  (a)  the  distribution  of  qualifications  among  market  entrants  (education  and  training  systems),  (b)  the 
effects  of  institutional  labour  market  systems,  encompassing  both  the  estimated  main  effects  and  the 
interaction  between  system  type  and  education  (labour  market  context),  (c)  aggregate  labour  market 
conditions,  and  (d)  the  extent  of  educational  expansion,  as  measured  by  the  proportion  of  tertiary, 
degrees  in  the  market.  In  the  first  column  of  the  table,  the  aggregate  prediction  at  average  context 
conditions  are  repeated  from  the  respective  tables  in  the  above  section.  The  last  column  provides  the 
empirical  aggregate  system  differences  at  otherwise  average  context  conditions,  expressed  as  the 
deviation  from  the  overall  average  in  the  first  column.  As  such,  the  well-known  advantage  of  OLM 
systems  with  respect  to  market  entrants’  unemployment  risks  is  expressed  as  a -.11  reduction  of  the 
overall  unemployment  rate  as  compared  to  the  EU  average  prediction  of  .20.  On  the  other  hand,  the 
Southern  countries  empirically  deviate  sharply  from  this  average  by  about  the  same  factor,  yet  in 
opposite  direction.  In  similar  ways,  empirical  system  differences  are  provided  for  the  different 
indicators  of  occupational  allocation,  which  also  reflect  results  reported  in  earlier  sections:  there  are 
only  minor  systemic  differences  in  terms  of  ISEI  status  scores,  while  OLM  systems  perform  markedly 
better  both  in  terms  of  a lower  proportion  of  lower-skilled  employment  among  market  entrants  (-.10 
reduction  from  the  average  prediction  of  .33)  and  a larger  proportion  of  young  people  in  professional 
employment  positions  (.09  increase  from  the  average  prediction  of  .21).  Notably  with  respect  to  the 
latter  dimension,  market  entrants  in  Southern  Europe  hold  such  advantageous  positions  to  a markedly 
smaller  extent  (-.08  decrease  from  an  average  of  .21).  In  most  cases,  ILM  systems  score  close  to  the 
EU  average. 

To  what  extent  can  these  empirical  differences  be  related  to  systemic  differences  in  the  role  of 
education  and  in  the  distribution  of  qualifications  among  school  leavers?  Columns  2-4  hold  the 
information  of  main  interest  here,  as  they  give  the  predicted  institutional  differences  to  these  factors, 
both  in  terms  of  their  single  and  combined  effects.  Columns  5 and  6 add  country  differences  in 
aggregate  unemployment  and  the  degree  of  educational  expansion  to  this,  providing  some  indications 
of  how  well  overall  differences  between  systems  can  be  reproduced  from  a selected  set  of  factors 
controlled  in  the  multivariate  models.  And  although  the  precise  figures  of  course  vary,  the  general 
conclusion  is  that  institutional  differences  related  to  market  entrants’  educational  resources  tend  to 
explain  a major  part  of  the  empirically  observed  differences  between  systems  on  all  four  indicators. 


This  can  be  more  easily  discussed  for  the  two  systems  deviating  clearly  from  the  EU  average,  the 
OLM  and  the  Southern  ones.  Taking  a look  at  the  decompositions  of  OLM  system  performance  first, 
institutional  differences  related  to  education  and  training  explain  about  two  thirds  of  the  favourable 
system  performance  in  terms  of  unemployment  (.072/.113),  all  of  the  (small)  advantage  in  terms  of 
status  attainment,  about  40%  of  the  favourable  performance  in  terms  of  lower-skilled  employment 
(.037/.10),  and  finally,  more  than  half  the  favourable  situation  with  respect  to  professional  employment 
outcomes  (.05/.089).  In  each  case,  the  favourable  situation  for  market  entrants  in  OLM  systems  is 
additionally  fostered  by  a currently  relatively  favourable  economic  situation,  but  this  effect  is  generally 
smaller  than  the  institutional  difference.  Considering  the  educational  effects  in  more  detail,  it  is  also 
apparent  that  the  advantage  of  OLM  systems  rests  on  both  a more  favourable  institutional  context  in 
labour  markets  and  a more  favourable  qualificational  background  of  market  entrants,  although  the 
precise  relation  varies  depending  on  the  indicator  chosen.  With  respect  to  unemployment,  both  effects 
are  of  equal  importance.  In  the  case  of  occupational  allocation,  it  is  mostly  a relatively  favourable 
qualificational  background  of  market  entrants  that  generates  the  system-level  advantage  in  terms  of 
lower-skilled  employment,  while  in  the  case  of  professional  employment,  there  is  evidence  of  a 
relatively  strong  role  of  favourable  institutional  conditions  in  the  labour  market  in  addition  to  the 
positive  effects  of  differences  in  market  entrants’  educational  background. 

The  situation,  interestingly,  is  quite  different  in  the  Southern  systems.  Here,  institutional  differences 
related  to  the  role  and  distribution  of  education  among  market  entrants  similarly  explain  a major  part  of 
the  observed  deviating  pattern,  in  all  cases  reproducing  or  even  overestimating  actual  system 
differences.  In  contrast  to  the  OLM  pattern,  the  relative  importance  of  each  component  in  bringing 
about  this  overall  result  is  clear-cut,  however.  Southern  Europe  deviates  in  terms  of  aggregate 
unemployment  because  the  institutional  context  of  labour  markets  works  markedly  against  young 
people  at  all  levels  of  education;  the  additional  contribution  of  a still  less  favourable  educational 
background  of  market  entrants  is  only  comparatively  small.  In  turn,  lower  occupational  outcomes  for 
market  entrants  are  mostly  related  to  the  relatively  unfavourable  educational  distribution  among 
market  entrants  rather  than  variation  in  the  role  of  education  for  occupational  attainment. 


ERIC 


5.31 


337 


Table  10  Institutional  Systems  and  Labour  Market  Entry:  Decomposing  System  Performance 


Average 

Prediction 

(a) 

Educational 

Systems 

Educational 

Distribution 

System  Differences 
(b)  (a+b)  (a+b+c) 

Labour  Educational  Education  + 

Market  Systems  + Aggregate 

Context  Labour  Unemploy- 

Market  ment  Rate 

Context 

(a+b+c+d) 
Education  + 
Unemploy- 
ment Rate  + 
% Tertiary 
Education 

Empirical 
Differences 
(at  average 
context 
conditions) 

Unemployment  Risks 

Probability  of  Unemployment 

ILM  Systems 

0.202 

+0.007 

-0.014 

-0.009 

+0.059 

+0.058 

+0.015 

OLM  Systems 

0.202 

-0.040 

-0.046 

-0.072 

-0.119 

-0.119 

-0.113 

Southern  Europe 

0.202 

+0.037 

+0.115 

+0.129 

+0.130 

+0.131 

+0.116 

Occupational  Allocation 

Status  Attainment  (ISEI  Score) 

ILM  Systems 

41.4 

+0.51 

-0.28 

+0.31 

+0.04 

-0.81 

-0.20 

OLM  Systems 

41.4 

+1.00 

+0.80 

+ 1.93 

+2.30 

+1.89 

+1.74 

Southern  Europe 

41.4 

-2.36 

-0.59 

-2.77 

-2.77 

-0.91 

-1.85 

Probability  of  Lower-Skilled  Employment 

ILM  Systems 

0.329 

-0.003 

+0.005 

+0.006 

+0.020 

+0.033 

+0.051 

OLM  Systems 

0.329 

-0.046 

-0.001 

-0.037 

-0.053 

-0.048 

-0.100 

Southern  Europe 

0.329 

+0.066 

+0.003 

+0.059 

+0.059 

+0.025 

+0.024 

Probability  of  Professional  Employment 

ILM  Systems 

0.208 

+0.017 

-0.012 

+0.003 

-0.012 

-0.024 

-0.019 

OLM  Systems 

0.208 

+0.022 

+0.033 

+0.050 

+0.069 

+0.064 

+0.089 

Southern  Europe 

0.208 

-0.064 

-0.018 

-0.080 

-0.080 

-0.060 

-0.076 

Notes:  Predicted  system  differences  compared  to  the  average  prediction:  all  predictions  are  based  on  the  final 
model  for  each  dependent  variable.  Cf.  Tables  1 , 3,  5,  and  7 for  further  details. 


6 Summary  and  Conclusions 

In  sum,  the  empirical  analyses  provided  reiterate  the  standard  result  of  the  crucial  role  of  education  for 
successful  integration  of  young  people  into  the  labour  market.  Even  in  this  early  stage  of  careers, 
education  provides  a highly  productive  resource  for  favourable  labour  market  attainment.  In  general, 
market  entrants  with  higher  levels  of  education  experience  lower  unemployment  risks  and  more 
favourable  occupational  allocation  in  their  initial  years  in  the  labour  force.  In  addition,  vocational 
specialization  and  notably  the  completion  of  vocational  training  in  an  apprenticeship  or  dual  system 
environment  markedly  reduce  unemployment  risks  as  compared  to  those  of  both  lower  secondary 
leavers  and  leavers  with  academic  upper  secondary  education.  These  results  hold  across  a sample  of 
twelve  European  countries  in  the  mid-1990s,  controlling  for  very  different  economic  context  conditions 
prevailing  in  these  countries  and  allowing  for  otherwise  unmeasured  heterogeneity  between  countries 
and  types  of  education. 


r ;•  5.32 


3.1,9 


Insofar  as  European  education  and  training  systems  produce  markedly  different  qualification  profiles 
among  market  entrants,  these  results  bear  important  implications  for  an  institutional  explanation  of 
cross-national  differences  in  labour  market  outcomes  for  those  entering  the  market.  As  education  and 
training  systems  vary  in  their  effectiveness  in  of  providing  young  people  with  productive  qualificational 
resources  valued  in  labour  markets,  a large  part  of  cross-national  differences  in  market  entrants’ 
unemployment  rates  or  initial  occupational  attainment  can  consistently  be  related  to  compositional 
differences  between  countries  in  the  distribution  of  qualifications  among  market  entrants.  Contrasting 
the  group  of  dual  system  type  countries  to  the  other  Northern  European  countries,  these  compositional 
differences  alone  account  for  about  50%  of  the  empirical  differential  in  terms  of  unemployment  rates 
and  still  around  one  third  of  the  differential  in  the  incidence  of  lower-skilled  employment.  Clearly,  the 
institutional  structure  of  education  and  training  systems  is  a most  powerful  factor  in  achieving  smooth 
transition  patterns  between  education  and  work  and  a quick  integration  of  young  people  into  working 
life.  Given  that  the  major  division  between  Northern  European  countries  occurs  in  terms  of  the  relative 
importance  of  vocational  training  at  the  upper  secondary  level  and,  associated  to  it,  in  the  extent  of 
dual  system  arrangements  for  training  provision,  it  is  clear  that  such  training  arrangements  constitute  a 
major  source  of  smoother  transition  patterns  in  dual  system  countries.  Additionally,  there  is  some' 
evidence  that  current  lower  levels  of  education  among  market  entrants  in  Southern  Europe  partially 
contribute  to  less  favourable  transition  outcomes,  although  this  applies  more  to  lower  aggregate  levels 
of  occupational  attainment  than  in  the  case  of  unemployment. 

In  assessing  these  results  on  the  importance  of  education  and  training  systems  for  explaining  cross- 
national differences  in  labour  market  entry,  it  should  be  emphasized  again  that  the  estimated  models 
provide  a substantial  advancement  beyond  available  evidence  as  they,  in  contrast  to  alternative 
studies,  are  derived  from  an  explicitly  microlevel  estimation  which  additionally  incorporates  an 
extensive  set  of  measured  and  unmeasured  macrolevel  controls  in  a simultaneous  estimation.  This 
implies  that  even  after  statistically  controlling  for  the  effects  of  varying  economic  context  conditions  on 
transition  outcomes,  fundamental  institutional  differences  operate  in  bringing  about  observed  cross- 
national differences.  This  not  only  includes  the  compositional  effects  discussed  above,  but  also  the 
impact  of  institutional  arrangements  in  labour  markets,  as  explored  by  a fairly  crude  distinction 
between  three  types  of  labour  market  systems  in  the  analyses.  According  to  the  results,  there  is 
considerable  evidence  that  rigid  institutional  conditions  in  the  Southern  European  labour  markets  work 
against  young  people  entering  the  market,  irrespective  of,  and  in  a sense  devaluing,  earlier 
investments  in  education  and  training.  Entering  the  labour  market  in  a Southern  European  context  is 
likely  to  imply  a much  extended  period  of  unemployment  before  securing  a first  job  as  compared  to  the 
Northern  European  experience,  even  after  controlling  for  education  and  economic  context  factors  (cf. 
also  Russell  and  O’Connell,  2000).  Of  course,  the  present  analysis  fails  to  identify  the  precise 
institutional  sources  for  this  pattern.  Nevertheless,  the  comparative  analysis  strongly  suggests 
avoiding  oversimplified  accounts  of  the  Southern  peculiarity.  The  lack  of  contract  regulations  allowing 
the  adjustment  of  job  rewards  and  initial  productivity  certainly  have  some  role  to  play  in  this  respect. 
Still,  there  is  the  intriguing  observation  that  it  is  the  highly  qualified  who  face  integration  problems 
rather  than  the  lowest  qualified  (as  in  Northern  Europe)  who  would  generally  be  expected  to  bear  the 
^■nemployment  costs  of  strong  employment  protection.  Any  explanation  based  on  the  effects  of  strict 

5.33 


339 


employment  protection  would  thus  have  to  come  to  terms  with  the  fact  that  the  lowest  qualified  do  not 
do  worse  in  Southern  Europe  as  compared  to  Northern  European  economies.  As  far  as  I can  see, 
there  are  two  potential  factors  that  might  explain  this  pattern.  The  first  argument  would  continue  to 
focus  on  the  detrimental  effects  of  strict  employment  protection  on  young  people’s  employment 
chances,  but  would  argue  that  family  networks  in  the  small-firm  based  Southern  economies  effectively 
provide  the  functional  substitute  which  generates  the  flexibility  necessary  to  integrate  less  qualified 
leavers.  If  this  were  the  case,  one  would  of  course  also  expect  different  selectivity  patterns  to  operate 
at  the  secondary  level  of  educational  systems.  Another  argument  consistent  with  the  observed 
patterns  would  be  to  focus  on  differences  in  supply-side  reactions  to  institutional  rigidities  in  the  labour 
market.  If  new  entrants  to  the  labour  market  intend  and,  indeed  are,  to  a larger  extent  enabled  to  fulfil 
their  occupational  aspirations  early  in  their  careers,  then  one  would  also  expect  a selection  of  relatively 
highly  educated  into  unemployment  (cf.  the  accounts  in  e.g.  Roberts  et  al.,  1999).  Obviously,  much 
richer  databases  than  the  Labour  Force  Surveys  are  needed  to  test  these  implications. 

At  the  same  time  as  institutional  rigidities  in  the  Southern  labour  markets  are  indicated,  the  analyses 
also  point  to  more  favourable  conditions  for  labour  market  entrants  in  OLM  type  labour  market 
contexts.  These  positive  effects  occur  in  terms  of  unemployment  risks,  but  also  with  respect  to 
avoiding  lower-skilled  employment  and  access  to  employment  in  the  professional  sector.  The  results 
are  consistent  with  arguments  related  to  the  operation  of  strong  occupational  labour  markets,  which 
tend  to  organize  worker  competition  around  skills  rather  than  experience  but  whose  viability  depends 
heavily  on  the  presence  of  a strong  vocational  training  system  (Marsden,  1990).  The  empirical  findings 
can  easily  be  read  as  indicating  a more  immediate  match  between  qualifications  and  employment 
positions  in  these  systems.  In  particular,  better  educated  leavers  have  a systematically  lower 
probability  of  starting  their  careers  in  lower  level  jobs  and  university  graduates  are  significantly  more 
likely  to  immediately  enter  the  professional  segment.  Even  more  compelling  is  the  evidence  with 
respect  to  unemployment.  In  the  OLM  economies,  it  is  vocational  training  (independent  of  its  source) 
which  provides  for  smoother  market  entry,  in  the  other  Northern  European  countries,  it  is  the  linkage  to 
a specific  employer  through  apprenticeship  contracts  or  other  dual  types  of  training  provision.  Adding 
up  the  institutional  effects  discussed  allows  us  to  explain  about  70%  of  the  favourable  performance  of 
the  set  of  OLM  systems  in  terms  of  market  entrants’  unemployment  risks,  and  still  50%  and  40%  of  the 
advantage  in  terms  of  early  access  to  professional  employment  and  the  incidence  of  lower-skilled 
employment,  respectively. 

A final  remark  about  the  role  of  education.  As  a whole,  the  stability  of  the  labour  market  value  of  the 
different  types  of  education  across  national  and  institutional  contexts,  as  established  in  the  analyses, 
suggests  that  we  view  qualifications  mainly  as  a productive  resource  of  market  entrants  in  securing  an 
adequate  start  to  working  life  rather  than  a mere  credentialist  screening  device.  Education  certainly 
bears  signals  about  potential  productivity  to  employers,  but  apparently  the  capacities  conveyed  by 
similar  types  of  education  are,  broadly  speaking,  also  recognized  and  rewarded  by  employers  in  fairly 
similar  ways.  This  should  not  preclude  a discussion  of  the  relative  merits  of  better  or  worse  examples 
of  specific  national  qualifications  and  their  features  - in  fact,  this  is  even  accounted  for  in  the  estimated 
models.  The  point  is  that,  in  general,  it  pays  off  for  market  entrants  to  invest  in  training  in  precisely  the 


ways  predicted  from  a very  simple  theoretical  framework  relating  smooth  labour  force  integration  to  the 
expected  training  costs  of  market  entrants,  irrespective  of  more  specific  institutional,  economic  or  other 
national  circumstances.  In  sum,  this  also  veritably  reinstates  education  and  the  ways  and  means  of 
training  provision  as  a core  instrument  of  policy'considerations,  notably  at  the  lower  levels  of  education 
where  policy  interventions  are  probably  much  more  called  for. 


References 

Allmendinger,  Jutta,  1989:  Educational  systems  and  labour  market  outcomes.  European  Sociological 
Review  5,  231-250. 

Ashton,  David  N.,  and  Jonny  Sung,  1992:  The  Determinants  of  Labour  Market  Transitions:  An 
Exploration  of  Contrasting  Approaches.  Work,  Employment  & Society  6,  1-21 . 

Ashton,  David  N.,  Malcolm  Maguire  and  Mark  Spilsbury,  1990:  Restructuring  the  Labour  Market.  The 
Implications  for  Youth.  Houndmills:  Macmillan. 

Becker,  Gary  S.,  1993  [1964]:  Human  Capital.  A Theoretical  and  Empirical  Analysis  with  Special 
Reference  to  Education.  3rd  edition.  Chicago:  University  of  Chicago  Press. 

Bernardi,  Fabrizio,  Richard  Layte,  Antonio  Schizzerotto  and  Sheila  Jacobs,  1999:  Who  exits 
unemployment?  Institutional  features,  individual  characteristics  and  chances  of  getting  a job.  A 
comparison  of  Britain  and  Italy.  Presentation  to  the  4th  European  Conference  of  Sociology. 
Amsterdam:  mimeo. 

Bills,  David  B.,  1988:  Educational  Credentials  and  Promotions:  Does  Schooling  Do  more  than  Get  You 
in  the  Door?  Sociology  of  Education  61 , 52-60. 

Blalock,  Hubert  M.,  1984:  Contextual-Effects  Models:  Theoretical  and  Methodological  Issues.  Annual 
Review  of  Sociology  10,  353-372. 

Blossfeld,  Hans-Peter,  1986:  Career  Opportunities  in  the  Federal  Republic  of  Germany:  a dynamic 
approach  to  the  study  of  life-course,  cohort,  and  period  effects.  European  Sociological  Review 
2,  208-225. 

Borghans,  Lex,  and  Loek  Groot,  1999:  Educational  presorting  and  occupational  segregation.  Labour 
Economics  6,  375-395. 

Bowlus,  Audra  J.,  1995:  Matching  Workers  and  Jobs:  Cyclical  Fluctuations  in  Match  Quality.  Journal  of 
Labor  Economics  13,  335-350. 

Braun,  Michael,  and  Walter  Muller,  1997:  Measurement  of  Education  in  Comparative  Research. 
Comparative  Social  Research  16,  163-201. 

Brauns,  Hildegard,  Markus  Gangl,  and  Stefani  Scherer,  1999:  Education  and  Unemployment:  Patterns 
of  Labour  Market  Entry  in  France,  the  United  Kingdom  and  West  Germany.  MZES  Working 
Paper  No.  6/1999.  Mannheim:  MZES. 


ERIC 


5.35 


341 


Brauns,  Hildegard,  Susanne  Steinmann,  Annick  Kieffer,  and  Catherine  Marry,  1999:  Does  Education 
matter?  France  and  Germany  in  Comparative  Perspective.  European  Sociological  Review  15, 
61-89. 


Brauns,  Hildegard,  Walter  Muller,  and  Susanne  Steinmann,  1998:  Education  and  Social  Class:  A 
Comparative  Study  on  the  Consequences  of  Educational  Expansion  in  Germany,  France,  the 
UK  and  Hungary.  Pp.  75-130  in:  Network  on  Transitions  in  Youth  (Eds.),  Transitions  in  Youth: 
Combating  Exclusion.  Dublin:  ESRI. 

Breen,  Richard,  and  John  H.  Goldthorpe,  1997:  Explaining  educational  differentials:  towards  a formal 
rational  action  theory.  Rationality  & Society  9,  275-305. 

Breen,  Richard,  Damian  F.  Hannan  and  Richard  O’Leary,  1995:  Returns  to  Education:  Taking  Account 
of  the  Employers’  Perception  and  Use  of  Educational  Credits.  European  Sociological  Review 
11,59-73. 


Bryk,  Anthony  S.,  and  Stephen  W.  Raudenbush,  1992:  Hierarchical  linear  models:  applications  and 
data  analysis  methods.  Newbury  Park:  Sage. 

DiPrete,  Thomas  A.  and  David  B.  Grusky,  1990a:  The  Multilevel  Analysis  of  Trends  with  Repeated 
Cross-Sectional  Data.  Sociological  Methodology  20,  337-368. 

DiPrete,  Thomas  A.  and  David  B.  Grusky,  1990b:  Structure  and  Trend  in  the  Process  of  Stratification 
for  American  Men  and  Women.  American  Journal  of  Sociology  96,  107-143. 

Dronkers,  Jaap,  1999:  The  Changing  Balance  between  Educational  Qualifications,  Occupations  and 
Gender  in  the  Netherlands  and  five  other  EU  Memberstates  between  the  70’s  and  90’s. 
Amsterdam:  SCO-Kohnstamm  Instituut. 

Eurostat,  1992:  Labour  Force  Survey.  Methods  and  definitions.  1992  series.  Luxembourg:  Eurostat. 

Eurostat,  1996:  The  European  Union  Labour  Force  Survey.  Methods  and  definitions  1996. 
Luxembourg:  Eurostat. 

Evans,  Phil,  1999:  Occupational  Downgrading  and  Upgrading  in  Britain.  Economica  66,  79-96. 

Eyraud,  Frangois,  David  Marsden,  and  Jean-Jacques  Silvestre,  1990:  Occupational  and  internal 
labour  markets  in  Britain  and  France.  International  Labour  Review  129,  501-517. 

Gangl,  Markus,  2000a:  European  Perspectives  on  Labour  Market  Entry:  A Matter  of  Institutional 
Linkages  between  Training  Systems  and  Labour  Markets?  CATEWE  Working  Paper. 
Mannheim:  mimeo. 

Gangl,  Markus,  2000b:  Changing  Labour  Markets  and  Early  Career  Outcomes:  Labour  Market  Entry  in 
Europe  over  the  Past  Decade.  CATEWE  Working  Paper.  Mannheim:  mimeo. 

Ganzeboom,  Harry  B.  G.,  and  Donald  Treiman,  1996:  Internationally  Comparable  Measures  of 
Occupational  Status  for  the  1988  International  Standard  Classification  of  Occupations.  Social 
Science  Research  25,  201-239. 

Ganzeboom,  Harry  B.  G.,  Paul  M.  De  Graaf,  and  Donald  Treiman,  1992:  A Standard  International 
Socio-Economic  Index  of  Occupational  Status.  Social  Science  Research  21,  1-56. 

Garonna,  Paolo  and  Paul  Ryan,  1989:  Le  travail  des  jeunes,  les  relations  professionnelles  et  les 
politiques  sociales  dans  les  economies  avancees.  Formation  Emploi  n°25,  janvier-mars. 

Goldstein,  Harvey,  1995:  Multilevel  Statistical  Models.  2nd  edition.  London:  Arnold. 

Greene,  William  H.,  1997:  Econometric  analysis.  3rd  edition.  Upper  Saddle  River:  Prentice  Hall. 


5.36 


Gregg,  Paul,  and  Alan  Manning,  1997:  Skill-biased  change,  unemployment  and  wage  inequality. 
European  Economic  Review41,  1173-1200. 

Grubb,  David  and  William  Wells,  1993:  Employment  Regulation  and  Patterns  of  Work  in  EC  Countries. 
OECD  Economic  Studies  21,  7-58. 

Hannan,  Damian,  et  al.,  1999:  A Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe  (CATEWE):  A Conceptual  Framework.  ESRI  Working  Paper  No.  137(a).  Dublin: 
Economic  and  Social  Research  Institute. 

Hunter,  Alfred  A.,  and  Jean  McKenzie  Leiper,  1993:  On  formal  education,  skills  and  earnings:  The  role 
of  educational  certificates  in  earnings  determination.  Canadian  Journal  of  Sociology  18,  21-42. 

International  Labour  Organisation,  1990a:  Statistical  Sources  and  Methods.  Volume  3:  Economically 
Active  Population,  Employment,  Unemployment  and  Hours  of  Work  (Household  Surveys).  2nd 
edition.  Geneva:  ILO. 

International  Labour  Organisation,  1990b:  ISCO-88  International  Standard  Classification  of 
Occupations.  Geneva: ILO 

Jobert,  Annette,  1997:  Employment,  education  and  training  in  Italy.  The  role  of  the  regions,  employers’ 
organisations  and  unions.  Pp.  208-223  in:  Jobert,  Annette,  Catherine  Marry,  Lucie  Tanguy, 
and  Helen  Rainbird  (Eds.),  Education  and  Work  in  Great  Britain,  Germany  and  Italy.  London:' 
Routledge. 

Jonsson,  Jan  O.,  1999:  Explaining  Sex  Differences  in  Educational  Choice.  An  Empirical  Assessment 
of  a Rational  Choice  Model.  European  Sociological  Review  15,  391-406. 

Judge,  George  G.  et  al.,  1985:  The  theory  and  practice  of  econometrics.  2nd  edition.  New  York:  Wiley. 

Kerckhoff,  Alan  C.,  1993:  Diverging  pathways:  social  structure  and  career  deflections.  Cambridge: 
Cambridge  University  Press. 

Kerckhoff,  Alan  C.,  1995:  Institutional  Arrangements  and  Stratification  Processes  in  Industrial 
Societies.  Annual  Review  of  Sociology  15,  323-347. 

Kerckhoff,  Alan  C.,  1996:  Building  Conceptual  and  Empirical  Bridges  between  Studies  of  Educational 
and  Labour  Force  Careers.  Pp.  37-56  in:  Alan  C.  Kerckhoff  (ed.),  Generating  Social 
Stratification:  Toward  a New  Research  Agenda.  Boulder:  Westview  Press. 

Kerckhoff,  Alan  C.,  forthcoming:  Transition  from  School  to  Work  in  Comparative  Perspective.  To 
appear  in  Maureen  T.  Hallinan  (ed.),  Handbook  of  Sociology  of  Education,  mimeo. 

Layard,  Richard,  Stephen  Nickell,  and  Richard  Jackman,  1991:  Unemployment.  Macroeconomic 
Performance  and  the  Labour  Market.  Oxford:  Oxford  University  Press. 

Longford,  Nicholas  T.,  1993:  Random  coefficient  models.  Oxford:  Clarendon  Press. 

Longford,  Nicholas  T.,  1995:  in  Gerhard  Arminger  et  al.  (Eds.),  Handbook  of  statistical  modeling  for  the 
social  and  behavioral  sciences.  New  York:  Plenum. 

Marsden,  David,  1986:  The  End  of  Economic  Man?  Custom  and  Competition  in  Labour  Markets. 
Brighton:  Wheatsheaf. 

Marsden,  David,  1990:  Institutions  and  Labour  Mobility:  Occupational  and  Internal  Labour  Markets  in 
Britain,  France,  Italy  and  West  Germany.  Pp.  414-438  in:  Renato  Brunetta  and  Carlo 
Dell’Aringa  (Eds.),  Labour  Relations  and  Economic  Performance.  Houndmills:  Macmillan. 

Marsden,  David,  and  Paul  Ryan,  1995:  Work,  labour  markets,  and  vocational  preparation:  Anglo- 
German  comparisons  of  training  in  intermediate  skills.  Pp.  67-79  in:  Leslie  Bash  and  Andy 
Green  (Eds.),  Youth,  Education  and  Work.  London:  Kogan  Page. 


ERIC 


5.37 


Maurice,  Marc,  Francois  Sellier  and  Jean-Jacques  Silvestre,  1986:  The  social  foundations  of  industrial 
power,  a comparison  of  France  and  Germany.  Cambridge:  MIT  Press. 

Muller,  Walter,  and  Maarten  Wolbers,  1999:  Educational  attainment  of  young  people  in  the  European 
Union:  cross-country  variation  of  trends  over  time.  CATEWE  Working  Paper.  Mannheim, 
Maastricht:  mimeo. 

Muller,  Walter,  and  Yossi  Shavit,  1998:  The  Institutional  Embeddedness  of  the  Stratification  Process: 
A Comparative  Study  of  Qualifications  and  Occupations  in  Thirteen  Countries.  Pp.  1-48  in: 
Yossi  Shavit  and  Walter  Muller  (Eds.),  From  School  to  Work.  A Comparative  Study  of 
Educational  Qualifications  and  Occupational  Destinations.  Oxford:  Clarendon  Press. 

Muller,  Walter,  Susanne  Steinmann  und  Reinhart  Schneider,  1997:  Bildung  in  Europa.  Pp.  177-246  in: 
Stefan  Hradil  and  Stefan  Immerfall  (Eds.),  Die  westeuropciischen  Gesellschaften  im  Vergleich. 
Opladen:  Leske  + Budrich. 

OECD,  1996:  Growing  into  Work:  Youth  and  the  Labour  Market  over  the  1980s  and  1990s. 
Employment  Outlook  58,  109-159. 

OECD,  1997:  Education  at  a Glance.  OECD  Indicators  1997.  Paris:  OECD. 

OECD,  1998:  Getting  started,  settling  in:  the  Transition  from  Education  to  the  Labour  Market. 
Employment  Outlook  60,  81-122. 

OECD,  1999:  Employment  Protection  and  Labour  Market  Performance.  Employment  Outlook  61, 47-. 

Penn,  Roger,  Michael  Rose  and  Jill  Rubery  (Eds.),  1994:  Skill  and  Occupational  Change.  Oxford: 
Oxford  University  Press. 

Polachek,  S.  W.,  and  W.  S.  Siebert,  1993:  The  economics  of  earnings.  Cambridge:  Cambridge 
University  Press. 

Roberts,  Ken  et  al.,  1999:  The  Young  Unemployed  and  the  Mediterraneanisation  of  Youth  Transition 
in  East-Central  Europe.  Paper  presented  to  the  1999  Annual  Meeting  of  the  Eu-ropean 
Research  Network  on  Transitions  in  Youth.  Oslo:  mimeo. 

Rosen,  Sherwin,  1972:  Learning  and  experience  in  the  labour  market.  The  Journal  of  Human 
Resources  7,  326-342. 

Rosenbaum,  James  E.,  Takehiko  Kariya,  Rick  Settersten,  and  Tony  Maier,  1990:  Market  and  Network 
Theories  of  the  Transition  from  High  School  to  Work:  Their  Application  to  Industrialized 
Societies.  Annual  Review  of  Sociology  16,  263-299. 

Ruiz-Quintanilla,  Segundo  A.  and  Rita  Claes,  1996:  Determinants  of  Underemployment  of  Young 
Adults:  A Multi-Country  Study.  Industrial  and  Labour  Relations  Review  49,  424-438. 

Russell,  Helen,  and  Philip  J.  O’Connell,  2000:  Getting  a Job  in  Europe:  The  Transition  from 
Unemployment  to  Employment  among  Young  People  in  Nine  European  Countries.  Pp.  133- 
152  in  Torild  Hammer  (ed.),  Transitions  and  Mobility  in  the  Youth  Labour  Market.  Proceedings 
of  the  1999  Workshop  of  the  European  Research  Network  on  Transitions  in  Youth.  Oslo: 
NOVA. 

Scherer,  Stefani,  1999:  Early  Career  Patterns.  A Comparison  of  Great  Britain  and  West  Germany. 
MZES  Working  Papers  No.  7.  Mannheim:  MZES. 

Shavit,  Yossi  and  Walter  Muller  (Eds.),  1998:  From  School  to  Work.  A Comparative  Study  of 
Educational  Qualifications  and  Occupational  Destinations.  Oxford:  Clarendon  Press. 

Shavit,  Yossi,  and  Walter  Muller,  2000:  Vocational  secondary  education.  Where  diversion  and  where 
safety  net?  European  Societies  2,  29-50. 


5.38 


Shavit,  Yossi,  and  Walter  Muller,  forthcoming:  Vocational  Secondary  Education,  Tracking  and  Social 
Stratification.  To  appear  in  Maureen  T.  Hallinan  (ed.),  Handbook  of  Sociology  of  Education, 
mimeo. 

Sorensen,  Aage  B.,  and  Arne  L.  Kalleberg,  1 981 : An  Outline  of  a Theory  of  the  Matching  of  Persons  to 
Jobs.  Pp.  49-74  in:  Ivar  Berg  (Ed.),  Sociological  Perspectives  on  Labor  Markets.  New  York: 
Academic  Press. 

Soskice,  David,  forthcoming:  Divergent  Production  Regimes.  Coordinated  and  Uncoordinated  Market 
Economies  in  the  1980s  and  1990s.  To  appear  in  Herbert  Kitschelt,  Peter  Lange,  Gary  Marks 
and  John  D.  Stephens,  Continuity  and  Change  in  Contemporary  Capitalism.  Cambridge: 
Cambridge  University  Press. 

Spence,  Michael,  1973.  Job  Market  Signaling.  Quarterly  Journal  of  Economics  87,  355-374. 

Spence,  Michael,  1981.  Signaling,  Screening,  and  Information.  In:  Rosen,  Sherwin  (ed.):  Studies  in 
Labour  Markets.  Chicago:  University  of  Chicago  Press. 

Spilerman,  Seymour  and  Lunde,  Tormod,  1991:  Features  of  Educational  Attainment  and  Job 
Promotion  prospects.  American  Journal  of  Sociology  97,  689-720. 

Storer,  P.,  1994:  Unemployment  dynamics  and  labour  market  tightness:  an  empirical  evaluation  of 
matching  function  models.  Journal  of  Applied  Econometrics  9,  389-419. 

Thurow,  Lester  C.,  1975:  Generating  Inequality.  Mechanisms  of  Distribution  in  the  U.S.  Economy.  New 
York:  Basic  Books. 

UNESCO,  1975:  International  Standard  Classification  of  Education  (ISCED).  Paris:  UNESCO. 

Van  der  Velden,  Rolf,  and  Maarten  Wolbers,  2000a:  The  Key  Power  of  Education.  Pp.  217-234  in 
Torild  Hammer  (ed.),  Transitions  and  Mobility  in  the  Youth  Labour  Market.  Proceedings  of  the 
1999  Workshop  of  the  European  Research  Network  on  Transitions  in  Youth.  Oslo:  NOVA. 

Van  der  Velden,  Rolf,  and  Maarten  Wolbers,  2000b:  The  integration  of  young  people  into  the  labour 
market  within  the  European  Union:  the  role  of  institutional  settings.  Maastricht:  mimeo. 


5.39 


345 


January  2001 


Prepared  as  part  of  the  TSER  project: 

Comparative  Analysis  of  Transitions  from  Education  to  Work  in 
Europe 


Changing  Labour  Markets  and  Early  Career 
Outcomes:  Labour  Market  Entry  in  Europe 
Over  the  Past  Decade 


Markus  Gangl 


Mannheim  Centre  for  European  Social  Research  (MZES) 

University  of  Mannheim 

P.O.  Box  10  34  62 

D-68131  Mannheim 

Germany 

E-mail:  Markus.Gangl@mzes.uni-mannheim.de 


ERiC 


346 


WORKING  PAPERS 


Abstract 


The  paper  addresses  the  issue  of  the  driving  forces  behind  recent  changes  in  labour  market  entry 
outcomes  in  Europe.  Based  on  data  for  12  European  countries  from  the  1988-1997  European 
Community  Labour  Force  Survey,  the  empirical  analyses  estimate  panel  data  models  to  assess  the 
effects  of  cyclical  changes  in  aggregate  economic  conditions,  changing  youth  cohort  sizes,  increasing 
educational  expansion  and  structural  changes  in  labour  demand  on  market  entrants  unemployment 
risks  and  occupational  allocation.  In  general,  it  is  found  that  unemployment  risks  have  closely  followed 
the  evolution  of  aggregate  economic  conditions  with  demographic  factors  having  only  a small  impact. 
Changes  in  occupational  allocation,  in  turn,  are  much  dependent  on  the  relative  evolution  of 
educational  expansion  and  professionalization  tendencies.  In  addition,  these  trends  do  not  affect  all 
leavers  evenly:  the  lowest  qualified  are  most  heavily  affected  by  cyclical  changes  in  economic 
conditions,  while  leavers  from  tertiary  level  education  have  been  more  strongly  affected  by  trends  of 
changing  occupational  attainment.  Most  discomforting,  however,  increasing  labour  force 
professionalization  is  found  to  contribute  to  increasing  labour  market  difficulties  among  the  lowest 
qualified. 


6.1 


O M to* 


1 Introduction’ 


A concern  for  the  changing  nature  of  education-to-work  transitions  is  widespread  in  current  studies  of 
early  labour  force  careers.  Issues  to  be  tackled  abound:  to  what  extent  do  recent  upsurges  in  youth 
unemployment  rates  reflect  a changing  world  economy  and  an  increasing  exclusion  of  the  least  skilled 
in  Western  economies?  Is  there  a general  trend  towards  the  devaluation  of  qualifications  in 
conjunction  with  a credentialist  rat  race  towards  achieving  ever  higher  levels  of  education,  leaving  the 
least  able  unfailingly  behind?  Do  traditional  and  relatively  stable  trajectories  into  working  life  dissolve 
in  favour  of  individualistic  and  erratic  job  hopping  patterns?  Despite  the  prevalence  of  such  rhetorics  of 
change  and  the  doubtless  relevance  of  changing  economic  context  conditions  for  understanding  and 
explaining  recent  changes  in  the  relation  between  education  and  early  career  trajectories,  the 
available  evidence  is  limited  at  best  (e.g.  Blossfeld,  1986,  1989;  Konietzka,  1998,  1999;  OECD,  1998 
for  exceptions),  not  least  due  to  the  lack  of  adequate  data  sources  to  study  mid-  or  even  longer-term 
labour  market  change.  This  is  unfortunate  for  any  structurally-minded  approach  to  the  study  of  labour 
force  entry  outcomes  as  more  easily  available  information  on  individual  characteristics  of  those 
entering  into  working  life  has  come  to  play  a more  prominent  role  in  empirical  analysis  than  theory 
would  necessarily  suggest.  More  awkwardly,  the  impacts  of  more  structural  context  conditions 
providing  opportunities  and  constraints  to  young  people  entering  the  labour  force  are  regularly  not 
even  addressed  in  the  interpretation  of  results,  even  if  changes  in  labour  market  outcomes  or  the 
stratification  processes  generating  them  are  diagnosed. 

This  reasoning  should,  however,  not  be  read  as  downplaying  the  role  of  individual  action,  resources 
and  stratification  processes  related  to  them  in  generating  labour  force  careers.  Rather,  the  argument  is 
that  observed  changes  in  transition  outcomes  should  first  and  foremost  be  understood  from  what  is 
called  period  effects  in  the  terminology  of  life  course  research,  i.e.  varying  structural  context  conditions 
which  exert  effects  on  all  current  participants  in  the  labour  market  as  they  reflect  the  changing  balance 
of  various  forces  determining  competition  and  allocation  on  the  market,  thus  framing  relevant 
behaviour  and  decisions  within  the  broader  structural  context.  Seen  this  way,  the  economic  context  in 
labour  markets,  broadly  understood,  is  evidently  a main  factor  shaping  education-to-work  transitions 
and  their  changing  nature  over  time.  As  labour  markets  tighten,  competition  for  jobs  aggravates, 
potentially  largely  at  the  expense  of  those  having  entered  the  market  only  recently  who  lack  work 
experience  and  longer-term  attachments  to  a specific  employer,  rendering  them  much  more 
vulnerable  to  job  separations  and  dismissals.  Similarly,  the  more  long-term  trends  of  educational 
expansion  and  the  changing  occupational  structure  might  impact  on  early  labour  market  outcomes  as 
patterns  of  occupational  allocation  change,  traditional  trajectories  of  entering  the  market  become  more 
and  more  obsolete,  new  competencies  are  required  by  employers  and  qualifications  become  in  part 


This  paper  has  been  prepared  as  part  of  the  CATEWE  research  project  (A  Comparative  Analysis  of 
Transitions  from  Education  to  Work  in  Europe),  funded  under  the  TSER  programme  of  the  European 
Commission  (Contract  SOEZ-CT97-2019-CATEWE).  Extensive  comments  and  suggestions  from  CATEWE 
members,  notably  Thomas  Couppie,  Damian  Hannan,  Mich6le  Mansuy,  Walter  Muller,  David  Raffe,  Emer 
Smyth  and  Maarten  Wolbers  are  gratefully  acknowledged. 


ERIC 


devaluated  by  their  more  widespread  availability.  Of  course,  each  of  these  trends  and  economic 
context  conditions  might  affect  different  leaver  groups  quite  differently,  thus  affecting  the  nature  of 
social  stratification  in  the  short  or  medium  run.  Finally,  the  impact  of  such  trends  may  or  may  not  be 
similar  across  all  countries  and  arguments  linking  institutional  arrangements  in  training  systems  and 
labour  markets  to  market  entrants  labour  force  outcomes  can  easily  be  extended  so  as  to  expect 
varying  degrees  of  institutional  sheltering  from  the  impacts  of  general  economic  trends. 

Drawing  on  a unique  database  covering  12  European  countries,  the  paper  seeks  to  address  these 
effects  of  changing  labour  market  conditions  on  early  career  outcomes  for  those  who  entered  the 
labour  market  between  the  late  1980s  and  the  late  1990s.  In  order  to  investigate  into  two  major 
aspects  of  recent  changes  in  transition  outcomes,  the  analyses  address  market  entrants’ 
unemployment  risks  and  their  occupational  allocation  in  the  early  career  stages.  Combining  repeated 
cross-sectional  microdata  for  a relatively  large  sample  of  countries  over  the  past  decade  moreover 
allows  us  to  estimate  panel  data  models  which  simultaneously  control  for  individual  and  structural 
effects  in  the  labour  market  attainment  process.  In  doing  so,  the  paper  represents  an  attempt  to 
decompose  successful  labour  market  entry  into  the  effects  of  individual  resources  and  characteristics,- 
notably  in  terms  of  qualifications,  and  the  impact  of  context  conditions,  including  both  institutional 
differences  between  countries  and  economic  differences  between  countries  and  over  time.  As  the 
effects  of  labour  market  conditions  on  transition  outcomes  have  - with  the  exception  of  local  labour 
market  conditions  (e.g.  Bynner,  1994;  Evans  and  Heinz,  1994;  Heinz,  1999)  - hardly  been 
systematically  addressed,  the  interest  of  the  current  paper  clearly  is  with  the  latter.  The  issues  of 
qualification  effects  and  their  linkages  to  specific  institutional  contexts  is  discussed  in  more  detail  in  a 
companion  paper  to  the  current  one,  which  applies  a similar  methodology  (cf.  Gangl,  2000b).  The 
following  section  discusses  the  theoretical  background  and  hypotheses  of  the  paper,  while  Section  3 
describes  the  database  and  research  design  applied  in  the  analyses.  The  empirical  results  are 
discussed  in  two  sections,  with  Section  4 providing  some  descriptive  results  on  trends  in  transition 
outcomes  over  the  past  decade  and  Section  5 containing  the  multivariate  analyses.  The  results  are 
summarized  in  a concluding  section. 


2 Economic  Context  and  Changing  Patterns  of  Labour  Market 
Entry 


2.1  Economic  Context  and  Labour  Market  Entry 


At  a most  general  level,  labour  market  outcomes  are  generated  from  the  stochastic  matching  between 
available  labour  supply  and  available  labour  demand,  representing  the  joint  outcome  from  worker 
choice  and  worker  opportunity  in  terms  of  vacancies  and  the  respective  evaluation  processes  of 
worker  and  job  characteristics  (cf.  Fine,  1998,  for  a review  of  labour  market  theory  and  notably  the 
more  structurally  oriented  labour  market  models  like  queueing  theory  [Thurow  1975;  Boudon  1974, 
1982]  and  more  recent  matching  models  [Kalleberg/Sorensen  1979;  Granovetter,  1981;  Eliason,  1995; 


O 

ERIC 


6.3 


349 


Logan,  1996]).  As  such,  labour  markets  are  best  understood  by  estimating  the  values  of  workers'  and 
employers’  resources  which  are  exchanged  in  the  market  as  well  as  the  nature  of  competition  and 
rules  of  allocation  that  prevail  in  the  market.  From  the  workers’  perspective,  the  labour  market 
exchange  is  ideally  an  exchange  of  qualifications  for  work  time  and  productivity  in  a given  job,  which  is 
rewarded  by  wages,  job  security,  promotion  opportunities,  fringe  benefits  and  the  like  (cf.  Akerlof, 
1982,  for  a more  extensive  discussion).  The  point  of  departure  for  discussing  the  impact  of  context 
effects  on  this  matching  and  exchange  relationship  is  to  note  that  these  mutual  evaluation  processes, 
and  consequently,  labour  market  outcomes  and  the  rewards  and  advantages  attached  to  valued 
resources  like  e.g.  education  or  professional  experience  are  fundamentally  relative  in  nature,  notably 
relative  in  terms  of  workers  and  employers  position  in  the  supply  and  demand  distributions  (Coleman, 
1991;  Sakamoto  and  Powers,  1995).  Economic  context  conditions  might  now  be  understood  as  a 
shorthand  for  describing  the  relative  balance  of  these  different  components  of  labour  market  matching 
processes,  and  hence,  describing  the  prevailing  conditions  of  competition,  allocation  and  matching. 

This  simplistic  framework  suggests  four  major  types  of  structural  trends  which  should  affect  the  level 
of  matching  activity  as  well  as  competition  and  allocation  patterns  on  the  labour  market,  namely 
contraction  or  expansion  forces  on  both  labour  supply  and  labour  demand  versus  distributional 
changes  within  the  available  supply  of  both  workers  and  jobs.  Changes  in  each  of  these  four  context 
conditions  can  be  expected  to  generate  a distinct  labour  market  response  as  labour  market  matching 
activities  adjust  accordingly.  Tacitly  assuming  non-immediate  adjustment  in  the  market  and  relatively 
fixed  factors  in  the  short  run,  these  different  effects  will  be  discussed  in  turn,  treating 
contraction/expansion  forces  first  and  turning  to  structural  shifts  in  the  supply  and  demand 
distributions  afterwards.  Contraction  and  expansion  forces  potentially  apply  to  either  labour  supply  or 
labour  demand;  consequently,  it  is  possible  to  distinguishe  between  supply-side  demographic 
pressures  and  demand-side  business  cycle  effects  here.1  Either  type  of  change  will  affect  labour 
market  outcomes  primarily  via  unemployment  risks  as  the  short-term  balance  between  supply  and 
demand  on  the  market  is  changed.  Moreover,  both  types  of  changes  should  have  offsetting  and 
counteracting  effects:  increasing  labour  supply  implies  a relative  oversupply  of  work  in  the  short  run, 
which  leads  to  individually  rising  probabilities  of  unemployment.  In  turn,  more  buoyant  labour  markets 
imply  increasing  relative  levels  of  labour  demand,  contributing  thus  to  generally  reduced 
unemployment  risks.  It  has  to  be  recognised,  however,  that  the  impact  of  these  expanding  or 
contracting  forces  is  not  necessarily  neutral  in  terms  of  observed  labour  market  stratification. 
Changing  conditions  of  competition  in  the  labour  market  should  come  mostly  at  the  expense  of  the 
relatively  least  competitive  individuals  (Coleman,  1991;  Sprengers,  1992).  As  competition  for  available 
jobs  tightens  - due  to  either  increasing  labour  supply  or  decreasing  labour  demand  - the  relative  value 
of  the  least  attractive  resources  in  the  market  should  decline  most  strongly.  Alternatively,  one  might 
also  argue  that  demand  contractions  occur  disproportionately  at  the  level  of  lower  skilled  employment 


Note  that  the  available  time  series  of  a decade  does  not  allow  a differentiation  between  short-term  cyclical  or 
longer-term  structural  expansions  and  contractions  of  labour  demand.  Hence,  I use  the  term  business  cycle  as 
a convenient  shorthand  for,  and  equivalent  to,  labour  demand  expansion  or  contraction. 


v ;•>  v 


6.4 


350 


(e.g.  Bowlus,  1995;  Storer,  1994).  In  consequence,  tightened  competition  is  expected  to  imply  higher 
unemployment  risks  for  the  most  recent  entrants  to  the  market  in  general  as  they  lack  the  work 
experience  of  more  adult  workers  and  to  the  least  qualified  in  particular  who  additionally  possess  the 
least  valuable  resources.  To  the  extent  that  downward  substitution  occurs  in  the  labour  market  as  a 
consequence  of  match  shortages,  contraction  trends  should  also  lead  to  negative  effects  on 
occupational  attainment.  Of  course,  the  respective  reverse  predictions  apply  in  the  case  of  supply 
contraction  or  demand  expansion  which  imply  loosened  competition  contexts  in  the  labour  market. 

A second  issue,  which  has  been  at  the  center  of  much  social  science  debate,  are  the  effects  of 
structural  changes  in  either  the  distribution  of  qualifications  or  jobs.  These  types  of  changes  will  be 
referred  to  as  educational  expansion  and  occupational  professionalization.  Both  types  of  changes 
have  been,  in  principle,  firmly  established  for  major  European  countries.  Educational  expansion  has 
been  traced  back  to  the  post-war  period  (e.g.  Muller  and  Wolbers  1999;  Miiller/Haun  1994;  Muller  et 
al.,  1997;  Brauns  1998),  with  most  recent  developments  involving  substantial  increases  in  the 
proportion  of  tertiary  level  graduates  in  the  population  and  among  market  entrants  in  most  countries. 
The  case  is  less  clear  with  respect  to  skill  changes  on  the  demand  side  (cf.  Penn  et  al.,  1994,  for  an- 
overview),  which  are  regularly  discussed  under  the  headings  of  upskilling,  professionalization  or  skill- 
biased  technological  change  (e.g.  Gallie  et  al.,  1998;  Berman  et  al.,  1999;  Gregg  and  Manning,  1997), 
all  implying  a shift  in  labour  demand  to  more  skill  intensive  labour.  It  is,  however,  unclear  so  far 
whether  observed  trends  represent  a unidirectional  and  secular  development  or  whether  they  should 
be  seen  as  part  of  an  ongoing  polarization  of  employment.  Due  to  data  limitations,  the  evolution  of 
lower-skilled  employment  cannot  be  explored  in  the  context  of  this  paper  and  the  issue  has  thus  to  be 
left  for  future  research.  In  addition,  a focus  on  professionalization  fits  well  with  the  empirical  fact  of 
educational  expansion  at  the  tertiary  level  as  a main  trend  for  labour  supply  over  the  past  decade. 

It  is  argued  here  that  these  two  types  of  structural  shifts,  in  contrast  to  the  expansion  and  contraction 
forces  in  the  market  discussed  earlier,  primarily  affect  occupational  allocation  and  occupational  returns 
to  education,  rather  than  unemployment  risks  (cf.  the  reasoning  in  Boudon  1974,  1982;  Collins  1979). 
Both  types  of  changes  affect  the  relative  availability  of  qualifications  and  particular  occupations,  and 
consequently,  the  relationships  between  qualifications  and  occupoational  allocation:  ceteris  paribus, 
increasing  availability  of  particular  qualifications  implies  decreasing  average  levels  of  occupational 
attainment  at  each  level  of  qualifications  as  processes  of  downward  substitution  in  the  system  will  be 
triggered.  Increased  professionalization  will,  in  turn,  increase  the  average  levels  of  occupational 
attainment  and,  notably,  improve  the  probability  of  securing  professional  employment  positions  as  the 
availability  of  such  high-skill  positions  increases.  Clearly,  both  types  of  changes  are  again  offsetting 
and  counteracting  factors,  so  that  the  net  outcome  is  a question  of  the  relative  development  of  both 
trends.  Existing  empirical  studies  tend  to  lend  partial  support  to  these  claims,  although  the  evidence  is 
much  contested  (e.g.  Brauns,  Muller,  Steinmann,  1998;  Handl,  1996;  van  der  Ploeg,  1994;  Dronkers, 
1999,  for  analyses  on  educational  expansion;  Barrett  et  al.,  1999;  Moll,  1992;  Parcel  and  Mueller, 
1989;  Sorensen  and  Blossfeld,  1989;  Nickell  and  Bell,  1995;  Gregg  and  Manning,  1997,  Gallie  et  al., 
1998,  for  studies  in  the  line  of  occupational  upskilling  and  technological  change).  In  contrast  to  most  of 


these  earlier  studies  that  are  restricted  to  a time  series  for  single  countries,  the  present  paper  holds 
the  promise  of  more  reliable  results  as  a sample  of  several  countries  can  be  used  in  the  analyses. 

Returning  to  the  theoretical  arguments,  it  should  be  recognized  that  these  types  of  changes  also 
potentially  affect  the  nature  of  stratification  in  early  labour  market  outcomes.  Given  the  nature  of 
changes  that  occurred  over  the  past  decade  in  Europe  (cf.  also  Section  4 below),  one  might  expect  an 
occupational  devaluation  of  credentials  due  to  educational  expansion  to  occur  mostly  at  the  tertiary 
level  of  education  as  in  most  European  countries  under  study  it  was  more  the  relation  between  upper 
secondary  and  tertiary  education  that  changed  rather  than  a further  decline  of  low  educational 
achievement  (cf.  Muller  and  Wolbers,  2000).  In  a similar  vein,  the  increasing  demand  for  professional 
employment  should  by  and  large  be  to  the  benefit  of  tertiary  level  educated  leavers  from  the 
educational  system,  so  that,  in  sum,  changes  in  labour  market  stratification  in  terms  of  occupational 
outcomes  should  mainly  result  from  net  changes  in  the  relative  positions  of  the  most  qualified,  while 
floor  effects  for  occupational  allocation  among  the  least  qualified  should  work  against  much  change 
there.  As  in  the  case  of  demographic  and  cyclical  effects  discussed  above,  one  could  finally  also 
argue  about  side  effects  of  educational  expansion  and  labour  market  professionalization  on  market; 
entrants’  risk  of  unemployment.  To  the  extent  that  skill  or  credential  requirements  increase, 
unemployment  risks  might  rise  as  well,  notably  of  course  among  the  least  qualified  so  that  the 
concentration  of  unemployment  among  this  group  should  increase. 


2.2  Institutional  Labour  Market  Contexts  and  Changing  Economic  Conditions 


Following  the  outline  of  some  general  mechanisms  linking  changing  economic  contexts  to  changing 
labour  force  outcomes  of  market  entrants,  comparative  analysis  not  only  provides  the  opportunity  to 
test  the  hypotheses  developed  above  but  also  an  approach  to  explore  whether  the  proposed 
relationships  generally  hold  or  whether  particular  institutional  arrangements  in  labour  markets  might 
actually  serve  to  attenuate  (or  exaggerate)  the  impact  of  changing  economic  context  conditions  on 
labour  market  entrants’  initial  career  experiences. 


0 

ERIC 


At  a fairly  general  level,  previous  research  on  education-to-work  transitions  has  regularly  contrasted 
two  polar  types  of  institutional  arrangements  in  European  labour  markets,  namely  occupational  labour 
market  systems  (OLM)  versus  internal  labour  market  systems  (ILM),  where  the  former  type  of  system 
is  said  to  operate  in  countries  with  strongly  vocationally  oriented  training  like  Germany,  Austria, 
Denmark  or  the  Netherlands  (Muller  and  Shavit,  1998;  Shavit  and  Muller,  2000,  forthcoming;  Hannan 
et  al.,  1999;  Allmendinger,  1989;  Kerckhoff,  1995,  forthcoming;  Marsden  1986,  1990;  Marsden  and 
Ryan,  1995;  Maurice  et  al,  1986).  Apart  from  the  discussion  of  the  role  of  education  in  either  type  of 
institutional  labour  market  context,  some  observers  have  also  commented  on  the  implications  of  the 
rigidities  present  in  the  ideal-typical  OLM  context  as  compared  to  more  flexible  labour  market 
structures  (Blossfeld,  1992;  Hannan  et  al.,  1999).  If  the  structure  of  external  markets  is  more  heavily 
segmented  along  occupational  lines  in  the  former  systems,  this  by  definition  implies  smaller 
substitution  potentials  across  submarkets,  which  might  slow  down  adaptation  processes  due  to 


•‘V- 


-sir  ««■ . 

r'* 


6.6 


ocr  0 


structural  change  (cf.  DiPrete  et  al.,  1997;  Blossfeld,  1992).  In  terms  of  the  concerns  raised  above, 
there  are  two  implications  with  respect  to  the  occupational  attainment  of  market  entrants  and  its 
responsiveness  to  particular  economic  changes:  first,  the  detrimental  effects  of  educational  expansion 
should  be  attenuated  in  OLM  systems  as  the  occupationalized  nature  of  job  competition  and  available 
skills  makes  downward  substitution  processes  less  attractive.  On  the  other  hand,  as  labour  force 
experience  is  a less  valued  resource  in  job  competition  in  an  OLM  context,  market  entrants  should  be 
able  to  benefit  to  a relatively  larger  degree  from  the  increasing  availability  of  attractive  professional 
employment  positions. 

The  issue  of  flexible  versus  more  regulated  institutional  labour  market  context  might,  moreover,  also 
effect  the  relationship  between  aggregate  economic  conditions  and  market  entrants'  risk  of 
unemployment.  If,  in  general,  more  flexible  employment  regulation  in  (youth)  labour  markets  allow  the 
consequences  of  cyclical  economic  up-  and  downswings  to  have  a stronger  influence  on  workers, 
unemployment  risks  should  cyclically  fluctuate  most  strongly  in  the  institutionally  most  flexible 
systems.  This  leads  one  to  expect  an  interaction  effect  in  terms  of  the  magnitude  of  cyclical  effects  on 
market  entrants’  unemployment  risks,  which  would  contrast  not  only  the  fairly  regulated  labour 
markets  of  OLM,  but  also  the  strictly  regulated  Southern  European  labour  markets  to  the  relatively 
flexible  ones  in  the  so-called  ILM  systems  (cf.  Grubb  and  Wells,  1993;  OECD,  1999;  Hartog  and 
Theeuwes,  1993  for  overviews).  The  empirical  analyses  will  now  begin  to  explore  these  and  the  above 
issues  in  more  detail. 


3 Data  and  Methodology 

Following  the  above  introduction,  this  paper  provides  a set  of  comparative  analyses  of  labour  market 
entry  in  the  countries  of  the  European  Union  in  the  late  1980s  and  early  to  mid-1990s,  focusing 
primarily  on  the  impact  of  changing  economic  contexts  on  market  entrants'  transition  outcomes.  In 
these  analyses,  data  for  twelve  European  countries  is  used,  drawing  on  the  1988-1997  European 
Community  Labour  Force  Surveys.2  This  database  provides  standardised,  cross-sectional  information 
on  labour  force  participation,  unemployment  and  various  aspects  of  employment  compiled  from  EU 
member  states'  national  Labour  Force  Surveys.  The  surveys  themselves  consist  of  large-scale 
national  samples  which  are  at  least  repeated  annually,  thus  providing  a unique  database  of  repeated 
cross-sectional  surveys  of  labour  market  behaviour  and  employment  issues  in  EU  countries  (cf. 


2 This  data  has  kindly  been  provided  by  EUROSTAT,  the  Statistical  Office  of  the  European  Union.  Of  course, 
EUROSTAT  is  neither  responsible  for  the  uses  made  of  the  data  nor  the  views  held  by  the  author.  The  twelve 
countries  chosen  for  analyses  are  Austria,  Belgium,  Denmark,  France,  Germany,  Greece,  Ireland,  Italy,  the 
Netherlands,  Portugal,  Spain  and  the  United  Kingdom.  Luxembourg  is  excluded  for  reasons  of  small  sample 
sizes  giving  unreliable  results,  while  Finland  and  Sweden  had  to  be  excluded  as  occupational  information  is 
only  provided  in  1997.  For  the  chosen  set  of  countries,  single  annual  observations  were  excluded  due  to 
breaks  in  (part  of)  the  time  series  or  other  unreliabilities,  mostly  related  to  substantial  changes  in  the  coverage 
of  current  training  activities  or  the  coding  of  educational  qualifications  (France  and  the  Netherlands  before 
1993  and  1992,  respectively;  Belgium,  Denmark  and  Ireland  in  1992).  Data  on  Austria  is  only  available  since 
1995,  when  that  country  joined  the  European  Union. 


0 


353 


EUROSTAT  1988,  1992,  1996,  for  extensive  details  on  the  database).  The  analyses  themselves 
address  the  responsiveness  of  labour  market  outcomes  for  market  entrants  from  four  educational 
levels  to  changing  economic  context  conditions  over  the  last  decade.  Market  entrants  are  defined  as 
those  individuals  who  left  the  education  and  training  system  within  the  last  five  years.3  Educational 
levels  are  distinguished  according  to  the  International  Standard  Classification  of  Education  (ISCED; 
UNESCO,  1975),  i.e.:  ISCED  levels  0-2  or  having  attained  no  more  than  lower  secondary 
qualifications,  ISCED  level  3 or  having  attained  upper  secondary  education,  ISCED  level  5 or  having 
attained  post-secondary  or  lower  tertiary  qualifications,  and  ISCED  levels  6-7  or  having  attained  full 
university  or  Ph.D.  degrees. 

Based  on  the  ECLFS  dataset,  unemployment  risks  and  occupational  attainment  are  analyzed  as  two 
main  aspects  of  early  labour  market  attainment.  With  respect  to  employment  and  unemployment,  the 
ECLFS  database  follows  standard  international  ILO  definitions  (cf.  ILO,  1990a),  while  occupations  are 
classified  according  to  the  2-digit  ISCO-68  codes  until  1991,  while  the  ISCO-88  COM  scheme  at  the  3- 
digit  level  is  applied  since  1992  (cf.  ILO  1990b,  EUROSTAT  1988,  1992,  1996).  In  the  current  paper,  a 
small  modification  to  the  ILO  concept  of  employment  is  applied:  in  an  analysis  of  early  labour  market 
experiences  and  the  transition  from  education  to  work,  it  appears  unwarranted  to  include  all  individuals 
having  worked  for  payment  or  profit  without  paying  attention  to  any  current  participation  in  education 
and  training,  which  might  actually  represent  their  primary  status.  Deviating  from  standard  ILO 
procedures,  all  individuals  participating  in  any  kind  of  initial  formal  education  and  training  are  therefore 
excluded  from  the  labour  force.4  After  all,  market  entrants  are  thus  defined  as  individuals  having 
(intermittently,  if  only)  completed  their  educational  careers.  For  this  group,  labour  market  outcomes 
are  investigated  in  terms  of  unemployment  risks  in  early  careers  and  initial  occupational  allocation,  the 
latter  being  measured  in  terms  of  occupational  status  and  the  attainment  of  professional  employment 
positions.  Below,  Overview  1 provides  more  specific  details  on  the  measurement  of  each  concept. 


3 As  individual  time  of  leaving  education  is  unavailable  in  the  database,  the  timing  of  market  entry  is  proxied  by 
typical  graduation  ages  for  the  different  levels  and  types  of  education  as  published  by  the  OECD  (1997). 

4 This  rather  open  and  imprecise  formulation  is  meant  to  include  participation  in  those  regulated  forms  of 
training  which  might  be  considered  in  some  way  as  ‘initial’,  while  excluding  those  types  of  part-time  education 
which  serve  to  enhance  individual  qualifications  while  already  working.  Attending  university,  upper  secondary 
schools  or  dual-system  types  of  training  would  be  examples  of  the  former;  attending  evening  schools  or  firm- 
based  training  courses  examples  for  the  latter.  Full  details  of  coding  are  available  from  the  author  on  request. 


6.8  354 


Overview  1 

Employment 

Modified  ILO  international  definition  of  employment  / labour  market 

& Labour  Force 

participation  (cf.  ILO  1990a):  participation  in  initial  training  considered 
as  primary  status 

Unemployment 

ILO  international  definition  of  unemployment  (cf.  ILO  1990a) 

Occupational  Status 

ISEI  international  socio-economic  index  score  (cf.  Ganzeboom  etal., 
1992;  1996)  matched  at  the  level  of  2-digit  ISCO-68  (1988-1991)  and 
3-digit  ISC088-C0M  (1992-1997)  occupational  detail 

Professional  Employment 

Professional  occupation  according  to  ISCO  classification  (cf.  ILO 
1990b): 

ISC068,  2-digit:  1-21,  35,  40,  50  (ECLFS  1988-1991) 

ISC088-C0M,  2-digit:  11, 12,  21-33  (ECLFS  1992-1997) 

(e.g.  teaching  and  scientific  professionals,  managers,  architects,  health 
professionals,  technicians) 

In  the  current  analyses,  these  individual-level  measures  are  complemented  by  a set  of  context  factor 
measures,  which  are  conceived  of  as  including  both  institutional  variables  and  labour  market  context 
factors.  As  introduced  in  the  theoretical  section  above,  the  variable  characterizing  institutional  labour 
market  contexts  amounts  to  a simple  differentiation  between  the  three  country  clusters  discussed 
earlier:  Austria,  Denmark,  Germany  and  the  Netherlands  are  classified  as  OLM  systems,  the  four 
Southern  European  countries  Greece,  Italy,  Portugal  and  Spain  are  grouped  together  in  another 
country  cluster,  while  the  remaining  Northern  European  countries  Belgium,  France,  Ireland,  and  the 
United  Kingdom  will  be  referred  to  as  ILM  systems.  More  interesting  to  the  current  paper  are  four 
macrolevel  measures  introduced,  indexing  the  four  types  of  structural  changes  in  economic  contexts 
identified  in  the  theoretical  reasoning  above.  Following  the  arguments  made  there,  the  analyses 
include  (a)  the  demographic  size  of  youth  cohorts  in  terms  of  the  youth-adult  ratio  in  the  labour  force, 
i.e.  the  ratio  of  market  entrants  to  experienced  workers  in  the  total  labour  force  aged  15-59,  following 
the  sample  specifications  detailed  above;  (b)  the  aggregate  unemployment  rate  in  the  total  labour 
force  aged  15-59,  indexing  aggregate  economic  conditions  and  business  cycle  fluctuations,  (c)  the 
extent  of  educational  expansion  as  captured  by  the  proportion  of  tertiary-level  - i.e.  ISCED  5-7  - 
qualified  individuals  in  the  total  labour  force,  and  (d)  the  extent  of  labour  market  professionalization  as 
measured  by  the  proportion  of  professional  employment  positions  (as  defined  in  Overview  1 ) among 
total  employment.5  All  measures  are  based  on  estimates  from  the  ECLFS  database  for  98  country 
level  observations,  i.e.  12  sample  countries  times  3-10  annual  observations.  The  analyses  themselves 
utilize  within-country  mean-centered  values  to  characterize  within-country  changes  in  economic 
contexts;  this  type  of  centering  allows  us  to  separate  genuine  trend  effects  from  level  effects  between 
countries. 


5 The  results  to  be  reported  below  are  qualitatively  unaffected  by  replacing  the  first  two  indicators  by 
employment  growth  and  changes  in  entry  cohort  sizes. 


O 

ERLC 


6.9 


355 


Summarizing  the  theoretical  discussion  in  operational  terms,  I expect  the  following  effects  for  each  of 
these  for  trend  factors: 

HI  Business  cycle  effects : Increasing  unemployment  rates  lead  to  increasing  risks  of  unemployment 
among  labour  market  entrants.  As  relatively  less  competitive  market  participants,  recent  entrants 
to  the  market  are  disproportionally  affected.  This  effect  also  interacts  negatively  with  the 
individual  level  of  qualifications,  so  that  the  lowest  qualified  are  affected  most  strongly.  Business 
cycle  effects  should,  moreover,  have  the  strongest  impact  on  initial  career  outcomes  in  the  most 
flexible  ILM  arrangements.  Effects  of  cyclically  increasing  unemployment  rates  on  occupational 
allocation  are,  finally,  weak  in  general  and  theoretically  indeterminate. 

H2  Demographic  effects:  Parallel  to  HI  above,  increasing  youth  cohort  sizes  are  expected  to  imply 
increasing  risks  of  unemployment,  again  potentially  disproportionately  affecting  the  least  qualified 
school  leavers.  Effects  on  occupational  allocation  are  weak  and  negative,  if  present. 

H3  Educational  expansion  effects:  Increasing  levels  of  tertiary  education  in  the  labour  force  lead  to  a 
devaluation  of  qualifications  in  terms  of  occupational  allocation.  Due  to  downward  substitution 
processes  triggered,  the  strongest  effects  will  be  on  occupational  outcomes  for  tertiary  level 
leavers.  In  the  context  of  occupational  labour  markets,  downward  substitution  tendencies  should 
be  attenuated.  Effects  of  educational  expansion  on  unemployment  risks  are  potentially  positive 
but  small. 

H4  Professionalization  effects:  Increasing  availability  of  professional  employment  positions  implies 
increasing  levels  of  occupational  attainment  among  market  entrants,  with  the  impact  again  being 
most  relevant  among  tertiary  level  leavers.  In  the  context  of  occupational  labour  markets,  market 
entrants  should  be  relatively  more  competitive  and  thus  benefit  relatively  more  from  the 
increasing  availability  of  professional  employment.  To  the  extent  that  skill  requirements  rise  in  the 
context  of  changing  labour  demand,  unemployment  risks  among  market  entrants  will  increase, 
notably  at  the  lower  levels  of  qualifications. 

Before  providing  tests  of  these  hypotheses  from  multivariate  models,  the  paper  will  first  present 
descriptive  evidence  on  trends  in  both  early  labour  market  outcomes  and  aggregate  economic  and 
labour  market  conditions  over  the  last  decade  in  the  twelve  EU  countries  under  study.  Their 
interrelations  and  the  impact  of  changes  in  aggregate  context  conditions  will  then  be  assessed  from 
comparative  micro-macro  models,  controlling  simultaneously  for  individual  resources,  institutional  and 
economic  context  factors,  and  unobserved  heterogeneity  between  countries  and  qualifications.  This 
modelling  strategy  follows  in  a straightforward  manner  from  the  repeated  cross-sectional  research 
design  of  the  database  used  (cf.  Blalock,  1984;  Judge  et  al.,  1985;  Greene,  1997;  DiPrete/Grusky 
1990a,  1990b)  and  the  time-series  cross-sectional  structure  of  the  data.  More  specifically,  I estimate  a 
set  of  panel  data  models  applying  Liang  and  Zeger's  (1986;  Diggle  et  al.,  1994)  GEE  estimation 
approach  which  extends  traditional  generalized  linear  modelling  based  on  cross-sectional  data  (cf. 
McCullagh  and  Nelders,  1989).  Effectively,  this  represents  a time-series  analysis  for  the  average 


labour  force  outcomes  of  market  entrants  from  48  educational  backgrounds  between  1988  and  1997. 
Compared  to  the  alternative  of  estimating  genuine  multi-level  models,  the  approach  might  be 
considered  the  more  robust  option  of  estimation  in  the  case  of  a fairly  small  number  of  aggregate  units 
of  analysis,  as  the  marginal  estimation  method  does  not  require  one  to  specify  and  estimate  mixtures 
of  distributions.  In  addition  to  the  time-varying  measures  of  economic  context  conditions,  all  estimated 
models  control  for  the  effects  of  levels  of  education  and  their  interaction  with  the  institutional  labour 
market  context  (cf.  Gangl,  2000b,  for  an  extensive  analysis  of  this  aspect)  and  allow  for  a first-order 
autoregressive  error  structure.  The  calculation  of  standard  errors  and  hypothesis  tests  is  adjusted  for 
the  clustering  on  48  educational  groups.  The  dichotomous  dependent  variables  of  unemployment  and 
attainment  of  professional  employment  are  modelled  by  specifying  a binomial  distribution  and  a logit 
link  function,  occupational  status  is  specified  to  follow  a normal  distribution  with  a logarithmic  link 
function. 


4 Trends  in  Transition  Outcomes  and  Economic  Contexts 
across  Europe 

As  a first  look  at  the  data,  Figure  1 below  gives  the  evolution  of  aggregate  labour  market  outcomes  for 
market  entrants  in  the  twelve  European  countries  between  1988  and  1997.  In  general,  the  trends 
exhibit  marked  cross-national  variation,  although  less  so  with  respect  to  changes  in  occupational 
allocation  outcomes.  With  respect  to  both  occupational  status  attainment  and  access  to  professional 
employment  positions,  the  situation  of  employed  market  entrants  improved  in  many  if  not  most 
European  countries  over  the  decade,  and  certainly  did  not  deteriorate  in  any  country.  Where 
occupational  allocation  outcomes  improved,  aggregate  gains  are  often  only  moderate,  but  in  some 
countries  considerably  increasing  levels  of  occupational  attainment  are  evident:  average  occupational 
status  attainment  increased  markedly  in  Spain  and  Portugal,  but  also  in  Germany  and  Belgium.  The 
results  for  access  to  professional  employment  are  similar.  Aggregate  trends  with  respect  to 
unemployment  are  less  easily  summarized.  For  most  countries,  there  is  evidence  of  strongly  cyclical 
patterns  in  the  evolution  of  market  entrants’  unemployment  rates,  with  falling  rates  up  to  the  early 
1990s  and  from  the  mid-  to  late  1990s,  while  unemployment  was  on  the  rise  over  the  early  1990s. 
There  is  substantial  variation,  however,  in  both  the  timing  and  the  extent  of  swings  observed.  In  most 
countries,  unemployment  rates  rose  by  around  5 percentage  points  in  the  early  1990s,  sometimes 
even  more,  with  a time  lag  of  about  three  years  between  Britain  and  Ireland  on  the  one  h